diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,11382 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.994219653179191,
+  "eval_steps": 500,
+  "global_step": 1620,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0030828516377649326,
+      "grad_norm": 6.770342510040142,
+      "learning_rate": 4.938271604938272e-07,
+      "loss": 1.0575,
+      "step": 1
+    },
+    {
+      "epoch": 0.006165703275529865,
+      "grad_norm": 6.728905738091506,
+      "learning_rate": 9.876543209876544e-07,
+      "loss": 1.0587,
+      "step": 2
+    },
+    {
+      "epoch": 0.009248554913294798,
+      "grad_norm": 6.675101590072679,
+      "learning_rate": 1.4814814814814815e-06,
+      "loss": 1.0534,
+      "step": 3
+    },
+    {
+      "epoch": 0.01233140655105973,
+      "grad_norm": 6.523315071031195,
+      "learning_rate": 1.9753086419753087e-06,
+      "loss": 1.0505,
+      "step": 4
+    },
+    {
+      "epoch": 0.015414258188824663,
+      "grad_norm": 6.109686845649886,
+      "learning_rate": 2.469135802469136e-06,
+      "loss": 1.0419,
+      "step": 5
+    },
+    {
+      "epoch": 0.018497109826589597,
+      "grad_norm": 4.769702242016632,
+      "learning_rate": 2.962962962962963e-06,
+      "loss": 1.0026,
+      "step": 6
+    },
+    {
+      "epoch": 0.02157996146435453,
+      "grad_norm": 4.372155317091635,
+      "learning_rate": 3.4567901234567904e-06,
+      "loss": 0.9988,
+      "step": 7
+    },
+    {
+      "epoch": 0.02466281310211946,
+      "grad_norm": 2.938055333553729,
+      "learning_rate": 3.9506172839506175e-06,
+      "loss": 0.967,
+      "step": 8
+    },
+    {
+      "epoch": 0.027745664739884393,
+      "grad_norm": 2.692200379748087,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.9602,
+      "step": 9
+    },
+    {
+      "epoch": 0.030828516377649325,
+      "grad_norm": 4.527289896935817,
+      "learning_rate": 4.938271604938272e-06,
+      "loss": 0.9497,
+      "step": 10
+    },
+    {
+      "epoch": 0.03391136801541426,
+      "grad_norm": 4.687306404911996,
+      "learning_rate": 5.432098765432099e-06,
+      "loss": 0.9548,
+      "step": 11
+    },
+    {
+      "epoch": 0.03699421965317919,
+      "grad_norm": 4.47710829915074,
+      "learning_rate": 5.925925925925926e-06,
+      "loss": 0.9431,
+      "step": 12
+    },
+    {
+      "epoch": 0.04007707129094412,
+      "grad_norm": 3.309898312513281,
+      "learning_rate": 6.419753086419753e-06,
+      "loss": 0.9022,
+      "step": 13
+    },
+    {
+      "epoch": 0.04315992292870906,
+      "grad_norm": 3.23269524121776,
+      "learning_rate": 6.913580246913581e-06,
+      "loss": 0.8804,
+      "step": 14
+    },
+    {
+      "epoch": 0.046242774566473986,
+      "grad_norm": 2.624187669592893,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.8637,
+      "step": 15
+    },
+    {
+      "epoch": 0.04932562620423892,
+      "grad_norm": 1.7038354361095256,
+      "learning_rate": 7.901234567901235e-06,
+      "loss": 0.8406,
+      "step": 16
+    },
+    {
+      "epoch": 0.05240847784200385,
+      "grad_norm": 1.2616042374824417,
+      "learning_rate": 8.395061728395062e-06,
+      "loss": 0.828,
+      "step": 17
+    },
+    {
+      "epoch": 0.055491329479768786,
+      "grad_norm": 1.6041822148570175,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 0.8077,
+      "step": 18
+    },
+    {
+      "epoch": 0.05857418111753372,
+      "grad_norm": 1.460418773609834,
+      "learning_rate": 9.382716049382717e-06,
+      "loss": 0.8,
+      "step": 19
+    },
+    {
+      "epoch": 0.06165703275529865,
+      "grad_norm": 1.0961373417740967,
+      "learning_rate": 9.876543209876543e-06,
+      "loss": 0.787,
+      "step": 20
+    },
+    {
+      "epoch": 0.06473988439306358,
+      "grad_norm": 1.0870323118053775,
+      "learning_rate": 1.037037037037037e-05,
+      "loss": 0.7734,
+      "step": 21
+    },
+    {
+      "epoch": 0.06782273603082851,
+      "grad_norm": 1.1132432560593115,
+      "learning_rate": 1.0864197530864198e-05,
+      "loss": 0.7634,
+      "step": 22
+    },
+    {
+      "epoch": 0.07090558766859345,
+      "grad_norm": 0.9201908286914825,
+      "learning_rate": 1.1358024691358025e-05,
+      "loss": 0.7532,
+      "step": 23
+    },
+    {
+      "epoch": 0.07398843930635839,
+      "grad_norm": 0.9373512896041561,
+      "learning_rate": 1.1851851851851852e-05,
+      "loss": 0.7551,
+      "step": 24
+    },
+    {
+      "epoch": 0.07707129094412331,
+      "grad_norm": 1.0444491432477614,
+      "learning_rate": 1.234567901234568e-05,
+      "loss": 0.7444,
+      "step": 25
+    },
+    {
+      "epoch": 0.08015414258188824,
+      "grad_norm": 0.9050511882090455,
+      "learning_rate": 1.2839506172839507e-05,
+      "loss": 0.7352,
+      "step": 26
+    },
+    {
+      "epoch": 0.08323699421965318,
+      "grad_norm": 0.9637872060692808,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.7333,
+      "step": 27
+    },
+    {
+      "epoch": 0.08631984585741811,
+      "grad_norm": 0.7660445344465261,
+      "learning_rate": 1.3827160493827162e-05,
+      "loss": 0.729,
+      "step": 28
+    },
+    {
+      "epoch": 0.08940269749518305,
+      "grad_norm": 0.8896905096851918,
+      "learning_rate": 1.4320987654320988e-05,
+      "loss": 0.7201,
+      "step": 29
+    },
+    {
+      "epoch": 0.09248554913294797,
+      "grad_norm": 0.6152980852860628,
+      "learning_rate": 1.4814814814814815e-05,
+      "loss": 0.7175,
+      "step": 30
+    },
+    {
+      "epoch": 0.09556840077071291,
+      "grad_norm": 0.5996149387942992,
+      "learning_rate": 1.5308641975308643e-05,
+      "loss": 0.7153,
+      "step": 31
+    },
+    {
+      "epoch": 0.09865125240847784,
+      "grad_norm": 0.5903944734308331,
+      "learning_rate": 1.580246913580247e-05,
+      "loss": 0.7043,
+      "step": 32
+    },
+    {
+      "epoch": 0.10173410404624278,
+      "grad_norm": 0.5316017360295336,
+      "learning_rate": 1.6296296296296297e-05,
+      "loss": 0.7016,
+      "step": 33
+    },
+    {
+      "epoch": 0.1048169556840077,
+      "grad_norm": 0.5314065222999873,
+      "learning_rate": 1.6790123456790123e-05,
+      "loss": 0.7019,
+      "step": 34
+    },
+    {
+      "epoch": 0.10789980732177264,
+      "grad_norm": 0.6467946203692391,
+      "learning_rate": 1.728395061728395e-05,
+      "loss": 0.6924,
+      "step": 35
+    },
+    {
+      "epoch": 0.11098265895953757,
+      "grad_norm": 0.49111482316088784,
+      "learning_rate": 1.7777777777777777e-05,
+      "loss": 0.6954,
+      "step": 36
+    },
+    {
+      "epoch": 0.11406551059730251,
+      "grad_norm": 0.6669723092878129,
+      "learning_rate": 1.8271604938271607e-05,
+      "loss": 0.6926,
+      "step": 37
+    },
+    {
+      "epoch": 0.11714836223506744,
+      "grad_norm": 0.5661110580326811,
+      "learning_rate": 1.8765432098765433e-05,
+      "loss": 0.6884,
+      "step": 38
+    },
+    {
+      "epoch": 0.12023121387283237,
+      "grad_norm": 0.6663410945644086,
+      "learning_rate": 1.925925925925926e-05,
+      "loss": 0.6848,
+      "step": 39
+    },
+    {
+      "epoch": 0.1233140655105973,
+      "grad_norm": 0.6194376397498395,
+      "learning_rate": 1.9753086419753087e-05,
+      "loss": 0.6773,
+      "step": 40
+    },
+    {
+      "epoch": 0.12639691714836224,
+      "grad_norm": 0.574490566591393,
+      "learning_rate": 2.0246913580246917e-05,
+      "loss": 0.6727,
+      "step": 41
+    },
+    {
+      "epoch": 0.12947976878612716,
+      "grad_norm": 1.008801728045859,
+      "learning_rate": 2.074074074074074e-05,
+      "loss": 0.6757,
+      "step": 42
+    },
+    {
+      "epoch": 0.1325626204238921,
+      "grad_norm": 1.1313553146882989,
+      "learning_rate": 2.123456790123457e-05,
+      "loss": 0.6731,
+      "step": 43
+    },
+    {
+      "epoch": 0.13564547206165703,
+      "grad_norm": 1.0830233919009031,
+      "learning_rate": 2.1728395061728397e-05,
+      "loss": 0.6699,
+      "step": 44
+    },
+    {
+      "epoch": 0.13872832369942195,
+      "grad_norm": 1.4150567592125278,
+      "learning_rate": 2.2222222222222227e-05,
+      "loss": 0.6698,
+      "step": 45
+    },
+    {
+      "epoch": 0.1418111753371869,
+      "grad_norm": 0.7997904788090316,
+      "learning_rate": 2.271604938271605e-05,
+      "loss": 0.6651,
+      "step": 46
+    },
+    {
+      "epoch": 0.14489402697495182,
+      "grad_norm": 1.0342019737867965,
+      "learning_rate": 2.320987654320988e-05,
+      "loss": 0.6702,
+      "step": 47
+    },
+    {
+      "epoch": 0.14797687861271677,
+      "grad_norm": 1.5617059029755216,
+      "learning_rate": 2.3703703703703703e-05,
+      "loss": 0.6663,
+      "step": 48
+    },
+    {
+      "epoch": 0.1510597302504817,
+      "grad_norm": 0.9229373350380571,
+      "learning_rate": 2.4197530864197533e-05,
+      "loss": 0.6599,
+      "step": 49
+    },
+    {
+      "epoch": 0.15414258188824662,
+      "grad_norm": 2.232377373378617,
+      "learning_rate": 2.469135802469136e-05,
+      "loss": 0.6674,
+      "step": 50
+    },
+    {
+      "epoch": 0.15722543352601157,
+      "grad_norm": 1.3217316863928317,
+      "learning_rate": 2.5185185185185187e-05,
+      "loss": 0.6596,
+      "step": 51
+    },
+    {
+      "epoch": 0.1603082851637765,
+      "grad_norm": 1.897886950243402,
+      "learning_rate": 2.5679012345679013e-05,
+      "loss": 0.662,
+      "step": 52
+    },
+    {
+      "epoch": 0.16339113680154144,
+      "grad_norm": 1.6751036096921057,
+      "learning_rate": 2.6172839506172843e-05,
+      "loss": 0.6596,
+      "step": 53
+    },
+    {
+      "epoch": 0.16647398843930636,
+      "grad_norm": 1.808418348177942,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.6578,
+      "step": 54
+    },
+    {
+      "epoch": 0.16955684007707128,
+      "grad_norm": 1.6423342079403511,
+      "learning_rate": 2.7160493827160497e-05,
+      "loss": 0.6498,
+      "step": 55
+    },
+    {
+      "epoch": 0.17263969171483623,
+      "grad_norm": 1.6027044537720114,
+      "learning_rate": 2.7654320987654323e-05,
+      "loss": 0.6483,
+      "step": 56
+    },
+    {
+      "epoch": 0.17572254335260115,
+      "grad_norm": 1.6105567211972927,
+      "learning_rate": 2.814814814814815e-05,
+      "loss": 0.648,
+      "step": 57
+    },
+    {
+      "epoch": 0.1788053949903661,
+      "grad_norm": 1.7443130429620175,
+      "learning_rate": 2.8641975308641977e-05,
+      "loss": 0.6471,
+      "step": 58
+    },
+    {
+      "epoch": 0.18188824662813102,
+      "grad_norm": 1.4613581354598622,
+      "learning_rate": 2.9135802469135807e-05,
+      "loss": 0.6445,
+      "step": 59
+    },
+    {
+      "epoch": 0.18497109826589594,
+      "grad_norm": 1.6877585039492473,
+      "learning_rate": 2.962962962962963e-05,
+      "loss": 0.6502,
+      "step": 60
+    },
+    {
+      "epoch": 0.1880539499036609,
+      "grad_norm": 1.3477366761501839,
+      "learning_rate": 3.012345679012346e-05,
+      "loss": 0.6484,
+      "step": 61
+    },
+    {
+      "epoch": 0.19113680154142582,
+      "grad_norm": 1.57024144235578,
+      "learning_rate": 3.061728395061729e-05,
+      "loss": 0.6413,
+      "step": 62
+    },
+    {
+      "epoch": 0.19421965317919074,
+      "grad_norm": 1.0956109706901844,
+      "learning_rate": 3.111111111111112e-05,
+      "loss": 0.6376,
+      "step": 63
+    },
+    {
+      "epoch": 0.1973025048169557,
+      "grad_norm": 1.849394419790938,
+      "learning_rate": 3.160493827160494e-05,
+      "loss": 0.6398,
+      "step": 64
+    },
+    {
+      "epoch": 0.2003853564547206,
+      "grad_norm": 1.7440700055149798,
+      "learning_rate": 3.209876543209877e-05,
+      "loss": 0.64,
+      "step": 65
+    },
+    {
+      "epoch": 0.20346820809248556,
+      "grad_norm": 1.7852836966630707,
+      "learning_rate": 3.259259259259259e-05,
+      "loss": 0.6404,
+      "step": 66
+    },
+    {
+      "epoch": 0.20655105973025048,
+      "grad_norm": 1.6206416138378672,
+      "learning_rate": 3.308641975308642e-05,
+      "loss": 0.6391,
+      "step": 67
+    },
+    {
+      "epoch": 0.2096339113680154,
+      "grad_norm": 1.7685800478629221,
+      "learning_rate": 3.3580246913580247e-05,
+      "loss": 0.6417,
+      "step": 68
+    },
+    {
+      "epoch": 0.21271676300578035,
+      "grad_norm": 1.4496909357747718,
+      "learning_rate": 3.4074074074074077e-05,
+      "loss": 0.6328,
+      "step": 69
+    },
+    {
+      "epoch": 0.21579961464354527,
+      "grad_norm": 1.7432586844040625,
+      "learning_rate": 3.45679012345679e-05,
+      "loss": 0.6382,
+      "step": 70
+    },
+    {
+      "epoch": 0.21888246628131022,
+      "grad_norm": 1.61616831606733,
+      "learning_rate": 3.506172839506173e-05,
+      "loss": 0.6327,
+      "step": 71
+    },
+    {
+      "epoch": 0.22196531791907514,
+      "grad_norm": 1.9981740027897072,
+      "learning_rate": 3.555555555555555e-05,
+      "loss": 0.6369,
+      "step": 72
+    },
+    {
+      "epoch": 0.22504816955684007,
+      "grad_norm": 1.4701021766825233,
+      "learning_rate": 3.604938271604938e-05,
+      "loss": 0.6353,
+      "step": 73
+    },
+    {
+      "epoch": 0.22813102119460502,
+      "grad_norm": 1.9800998540295955,
+      "learning_rate": 3.654320987654321e-05,
+      "loss": 0.6329,
+      "step": 74
+    },
+    {
+      "epoch": 0.23121387283236994,
+      "grad_norm": 1.6478748827771714,
+      "learning_rate": 3.703703703703704e-05,
+      "loss": 0.6337,
+      "step": 75
+    },
+    {
+      "epoch": 0.2342967244701349,
+      "grad_norm": 1.8009899393758098,
+      "learning_rate": 3.7530864197530867e-05,
+      "loss": 0.6341,
+      "step": 76
+    },
+    {
+      "epoch": 0.2373795761078998,
+      "grad_norm": 1.802410608294057,
+      "learning_rate": 3.80246913580247e-05,
+      "loss": 0.6278,
+      "step": 77
+    },
+    {
+      "epoch": 0.24046242774566473,
+      "grad_norm": 1.3790254781071283,
+      "learning_rate": 3.851851851851852e-05,
+      "loss": 0.6249,
+      "step": 78
+    },
+    {
+      "epoch": 0.24354527938342968,
+      "grad_norm": 1.2291141825891714,
+      "learning_rate": 3.901234567901235e-05,
+      "loss": 0.633,
+      "step": 79
+    },
+    {
+      "epoch": 0.2466281310211946,
+      "grad_norm": 2.670730627521506,
+      "learning_rate": 3.950617283950617e-05,
+      "loss": 0.6415,
+      "step": 80
+    },
+    {
+      "epoch": 0.24971098265895952,
+      "grad_norm": 1.963981051720273,
+      "learning_rate": 4e-05,
+      "loss": 0.6339,
+      "step": 81
+    },
+    {
+      "epoch": 0.2527938342967245,
+      "grad_norm": 2.240860253391614,
+      "learning_rate": 4.049382716049383e-05,
+      "loss": 0.631,
+      "step": 82
+    },
+    {
+      "epoch": 0.2558766859344894,
+      "grad_norm": 2.6342207736042718,
+      "learning_rate": 4.098765432098766e-05,
+      "loss": 0.6349,
+      "step": 83
+    },
+    {
+      "epoch": 0.2589595375722543,
+      "grad_norm": 2.3208295583865346,
+      "learning_rate": 4.148148148148148e-05,
+      "loss": 0.6325,
+      "step": 84
+    },
+    {
+      "epoch": 0.26204238921001927,
+      "grad_norm": 2.050250484741221,
+      "learning_rate": 4.197530864197531e-05,
+      "loss": 0.6249,
+      "step": 85
+    },
+    {
+      "epoch": 0.2651252408477842,
+      "grad_norm": 2.337988226514687,
+      "learning_rate": 4.246913580246914e-05,
+      "loss": 0.6257,
+      "step": 86
+    },
+    {
+      "epoch": 0.2682080924855491,
+      "grad_norm": 1.2947394723319468,
+      "learning_rate": 4.296296296296297e-05,
+      "loss": 0.6276,
+      "step": 87
+    },
+    {
+      "epoch": 0.27129094412331406,
+      "grad_norm": 2.7291506969796817,
+      "learning_rate": 4.345679012345679e-05,
+      "loss": 0.6379,
+      "step": 88
+    },
+    {
+      "epoch": 0.274373795761079,
+      "grad_norm": 1.2324510488923008,
+      "learning_rate": 4.395061728395062e-05,
+      "loss": 0.6287,
+      "step": 89
+    },
+    {
+      "epoch": 0.2774566473988439,
+      "grad_norm": 4.370230321235183,
+      "learning_rate": 4.444444444444445e-05,
+      "loss": 0.661,
+      "step": 90
+    },
+    {
+      "epoch": 0.28053949903660885,
+      "grad_norm": 4.105567150207915,
+      "learning_rate": 4.493827160493828e-05,
+      "loss": 0.6783,
+      "step": 91
+    },
+    {
+      "epoch": 0.2836223506743738,
+      "grad_norm": 2.290535632598349,
+      "learning_rate": 4.54320987654321e-05,
+      "loss": 0.6423,
+      "step": 92
+    },
+    {
+      "epoch": 0.28670520231213875,
+      "grad_norm": 2.5986674923402027,
+      "learning_rate": 4.592592592592593e-05,
+      "loss": 0.637,
+      "step": 93
+    },
+    {
+      "epoch": 0.28978805394990365,
+      "grad_norm": 2.3015863180813976,
+      "learning_rate": 4.641975308641976e-05,
+      "loss": 0.6467,
+      "step": 94
+    },
+    {
+      "epoch": 0.2928709055876686,
+      "grad_norm": 2.5901990912421615,
+      "learning_rate": 4.691358024691359e-05,
+      "loss": 0.6374,
+      "step": 95
+    },
+    {
+      "epoch": 0.29595375722543354,
+      "grad_norm": 1.9655351020554899,
+      "learning_rate": 4.7407407407407407e-05,
+      "loss": 0.6303,
+      "step": 96
+    },
+    {
+      "epoch": 0.29903660886319844,
+      "grad_norm": 1.9527538197135768,
+      "learning_rate": 4.7901234567901237e-05,
+      "loss": 0.6337,
+      "step": 97
+    },
+    {
+      "epoch": 0.3021194605009634,
+      "grad_norm": 1.924365116454199,
+      "learning_rate": 4.8395061728395067e-05,
+      "loss": 0.6291,
+      "step": 98
+    },
+    {
+      "epoch": 0.30520231213872834,
+      "grad_norm": 1.4423490571535786,
+      "learning_rate": 4.88888888888889e-05,
+      "loss": 0.6316,
+      "step": 99
+    },
+    {
+      "epoch": 0.30828516377649323,
+      "grad_norm": 2.1757066725433174,
+      "learning_rate": 4.938271604938272e-05,
+      "loss": 0.6292,
+      "step": 100
+    },
+    {
+      "epoch": 0.3113680154142582,
+      "grad_norm": 1.8736458639020233,
+      "learning_rate": 4.987654320987655e-05,
+      "loss": 0.6277,
+      "step": 101
+    },
+    {
+      "epoch": 0.31445086705202313,
+      "grad_norm": 1.6129503412577801,
+      "learning_rate": 5.037037037037037e-05,
+      "loss": 0.6273,
+      "step": 102
+    },
+    {
+      "epoch": 0.3175337186897881,
+      "grad_norm": 1.6400157484048863,
+      "learning_rate": 5.08641975308642e-05,
+      "loss": 0.6235,
+      "step": 103
+    },
+    {
+      "epoch": 0.320616570327553,
+      "grad_norm": 1.4124420341435686,
+      "learning_rate": 5.1358024691358027e-05,
+      "loss": 0.6296,
+      "step": 104
+    },
+    {
+      "epoch": 0.3236994219653179,
+      "grad_norm": 2.013438318363578,
+      "learning_rate": 5.1851851851851857e-05,
+      "loss": 0.6245,
+      "step": 105
+    },
+    {
+      "epoch": 0.3267822736030829,
+      "grad_norm": 1.9515082565056439,
+      "learning_rate": 5.234567901234569e-05,
+      "loss": 0.6251,
+      "step": 106
+    },
+    {
+      "epoch": 0.32986512524084777,
+      "grad_norm": 1.6735890290722966,
+      "learning_rate": 5.283950617283952e-05,
+      "loss": 0.6282,
+      "step": 107
+    },
+    {
+      "epoch": 0.3329479768786127,
+      "grad_norm": 1.9553717360583245,
+      "learning_rate": 5.333333333333333e-05,
+      "loss": 0.6245,
+      "step": 108
+    },
+    {
+      "epoch": 0.33603082851637767,
+      "grad_norm": 1.971490342069946,
+      "learning_rate": 5.382716049382716e-05,
+      "loss": 0.6178,
+      "step": 109
+    },
+    {
+      "epoch": 0.33911368015414256,
+      "grad_norm": 1.1738997570110703,
+      "learning_rate": 5.432098765432099e-05,
+      "loss": 0.6166,
+      "step": 110
+    },
+    {
+      "epoch": 0.3421965317919075,
+      "grad_norm": 2.9216989410368552,
+      "learning_rate": 5.481481481481482e-05,
+      "loss": 0.6274,
+      "step": 111
+    },
+    {
+      "epoch": 0.34527938342967246,
+      "grad_norm": 1.6405517346090435,
+      "learning_rate": 5.5308641975308647e-05,
+      "loss": 0.6239,
+      "step": 112
+    },
+    {
+      "epoch": 0.34836223506743735,
+      "grad_norm": 2.610418743905785,
+      "learning_rate": 5.580246913580247e-05,
+      "loss": 0.6335,
+      "step": 113
+    },
+    {
+      "epoch": 0.3514450867052023,
+      "grad_norm": 1.9525696348415542,
+      "learning_rate": 5.62962962962963e-05,
+      "loss": 0.6243,
+      "step": 114
+    },
+    {
+      "epoch": 0.35452793834296725,
+      "grad_norm": 2.0062636327574364,
+      "learning_rate": 5.679012345679013e-05,
+      "loss": 0.6291,
+      "step": 115
+    },
+    {
+      "epoch": 0.3576107899807322,
+      "grad_norm": 2.210750819858955,
+      "learning_rate": 5.728395061728395e-05,
+      "loss": 0.6285,
+      "step": 116
+    },
+    {
+      "epoch": 0.3606936416184971,
+      "grad_norm": 1.5402278190413057,
+      "learning_rate": 5.777777777777778e-05,
+      "loss": 0.6137,
+      "step": 117
+    },
+    {
+      "epoch": 0.36377649325626205,
+      "grad_norm": 2.4411415059275234,
+      "learning_rate": 5.827160493827161e-05,
+      "loss": 0.6248,
+      "step": 118
+    },
+    {
+      "epoch": 0.366859344894027,
+      "grad_norm": 1.4590091197032318,
+      "learning_rate": 5.876543209876544e-05,
+      "loss": 0.6201,
+      "step": 119
+    },
+    {
+      "epoch": 0.3699421965317919,
+      "grad_norm": 2.9813084309787574,
+      "learning_rate": 5.925925925925926e-05,
+      "loss": 0.6265,
+      "step": 120
+    },
+    {
+      "epoch": 0.37302504816955684,
+      "grad_norm": 2.241811235940394,
+      "learning_rate": 5.975308641975309e-05,
+      "loss": 0.6324,
+      "step": 121
+    },
+    {
+      "epoch": 0.3761078998073218,
+      "grad_norm": 2.267511302500656,
+      "learning_rate": 6.024691358024692e-05,
+      "loss": 0.6199,
+      "step": 122
+    },
+    {
+      "epoch": 0.3791907514450867,
+      "grad_norm": 2.015052931345159,
+      "learning_rate": 6.074074074074075e-05,
+      "loss": 0.6243,
+      "step": 123
+    },
+    {
+      "epoch": 0.38227360308285163,
+      "grad_norm": 1.943140665848731,
+      "learning_rate": 6.123456790123457e-05,
+      "loss": 0.621,
+      "step": 124
+    },
+    {
+      "epoch": 0.3853564547206166,
+      "grad_norm": 1.431350651784121,
+      "learning_rate": 6.17283950617284e-05,
+      "loss": 0.616,
+      "step": 125
+    },
+    {
+      "epoch": 0.3884393063583815,
+      "grad_norm": 2.0972998498743727,
+      "learning_rate": 6.222222222222223e-05,
+      "loss": 0.6227,
+      "step": 126
+    },
+    {
+      "epoch": 0.3915221579961464,
+      "grad_norm": 1.51053050639635,
+      "learning_rate": 6.271604938271606e-05,
+      "loss": 0.6208,
+      "step": 127
+    },
+    {
+      "epoch": 0.3946050096339114,
+      "grad_norm": 1.9909971477846713,
+      "learning_rate": 6.320987654320988e-05,
+      "loss": 0.613,
+      "step": 128
+    },
+    {
+      "epoch": 0.3976878612716763,
+      "grad_norm": 1.8273995721736398,
+      "learning_rate": 6.37037037037037e-05,
+      "loss": 0.617,
+      "step": 129
+    },
+    {
+      "epoch": 0.4007707129094412,
+      "grad_norm": 1.78165088558862,
+      "learning_rate": 6.419753086419754e-05,
+      "loss": 0.6161,
+      "step": 130
+    },
+    {
+      "epoch": 0.40385356454720617,
+      "grad_norm": 1.4335996844839054,
+      "learning_rate": 6.469135802469136e-05,
+      "loss": 0.6136,
+      "step": 131
+    },
+    {
+      "epoch": 0.4069364161849711,
+      "grad_norm": 1.790704705499662,
+      "learning_rate": 6.518518518518519e-05,
+      "loss": 0.6175,
+      "step": 132
+    },
+    {
+      "epoch": 0.410019267822736,
+      "grad_norm": 2.7954648767878614,
+      "learning_rate": 6.567901234567901e-05,
+      "loss": 0.6193,
+      "step": 133
+    },
+    {
+      "epoch": 0.41310211946050096,
+      "grad_norm": 1.189493692128073,
+      "learning_rate": 6.617283950617285e-05,
+      "loss": 0.6139,
+      "step": 134
+    },
+    {
+      "epoch": 0.4161849710982659,
+      "grad_norm": 2.7097095478470203,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 0.6432,
+      "step": 135
+    },
+    {
+      "epoch": 0.4192678227360308,
+      "grad_norm": 2.3706670733859867,
+      "learning_rate": 6.716049382716049e-05,
+      "loss": 0.622,
+      "step": 136
+    },
+    {
+      "epoch": 0.42235067437379575,
+      "grad_norm": 2.048042684280123,
+      "learning_rate": 6.765432098765433e-05,
+      "loss": 0.6314,
+      "step": 137
+    },
+    {
+      "epoch": 0.4254335260115607,
+      "grad_norm": 2.5636816467256787,
+      "learning_rate": 6.814814814814815e-05,
+      "loss": 0.6196,
+      "step": 138
+    },
+    {
+      "epoch": 0.42851637764932565,
+      "grad_norm": 1.9434364249998188,
+      "learning_rate": 6.864197530864199e-05,
+      "loss": 0.6252,
+      "step": 139
+    },
+    {
+      "epoch": 0.43159922928709055,
+      "grad_norm": 1.9573788625868929,
+      "learning_rate": 6.91358024691358e-05,
+      "loss": 0.619,
+      "step": 140
+    },
+    {
+      "epoch": 0.4346820809248555,
+      "grad_norm": 2.5321167605720136,
+      "learning_rate": 6.962962962962964e-05,
+      "loss": 0.6156,
+      "step": 141
+    },
+    {
+      "epoch": 0.43776493256262045,
+      "grad_norm": 1.4532436044075718,
+      "learning_rate": 7.012345679012346e-05,
+      "loss": 0.6167,
+      "step": 142
+    },
+    {
+      "epoch": 0.44084778420038534,
+      "grad_norm": 3.236415792135685,
+      "learning_rate": 7.06172839506173e-05,
+      "loss": 0.6142,
+      "step": 143
+    },
+    {
+      "epoch": 0.4439306358381503,
+      "grad_norm": 2.1157711195082363,
+      "learning_rate": 7.11111111111111e-05,
+      "loss": 0.6197,
+      "step": 144
+    },
+    {
+      "epoch": 0.44701348747591524,
+      "grad_norm": 3.1116355261096342,
+      "learning_rate": 7.160493827160494e-05,
+      "loss": 0.6167,
+      "step": 145
+    },
+    {
+      "epoch": 0.45009633911368013,
+      "grad_norm": 2.87114996778744,
+      "learning_rate": 7.209876543209877e-05,
+      "loss": 0.6189,
+      "step": 146
+    },
+    {
+      "epoch": 0.4531791907514451,
+      "grad_norm": 2.2936736836468556,
+      "learning_rate": 7.25925925925926e-05,
+      "loss": 0.6155,
+      "step": 147
+    },
+    {
+      "epoch": 0.45626204238921003,
+      "grad_norm": 1.9809595695858282,
+      "learning_rate": 7.308641975308643e-05,
+      "loss": 0.609,
+      "step": 148
+    },
+    {
+      "epoch": 0.4593448940269749,
+      "grad_norm": 1.787807991145882,
+      "learning_rate": 7.358024691358025e-05,
+      "loss": 0.61,
+      "step": 149
+    },
+    {
+      "epoch": 0.4624277456647399,
+      "grad_norm": 1.4507198750776926,
+      "learning_rate": 7.407407407407409e-05,
+      "loss": 0.605,
+      "step": 150
+    },
+    {
+      "epoch": 0.4655105973025048,
+      "grad_norm": 1.5130687282460713,
+      "learning_rate": 7.456790123456791e-05,
+      "loss": 0.6078,
+      "step": 151
+    },
+    {
+      "epoch": 0.4685934489402698,
+      "grad_norm": 1.2754368231242734,
+      "learning_rate": 7.506172839506173e-05,
+      "loss": 0.6058,
+      "step": 152
+    },
+    {
+      "epoch": 0.47167630057803467,
+      "grad_norm": 2.775374882049122,
+      "learning_rate": 7.555555555555556e-05,
+      "loss": 0.6159,
+      "step": 153
+    },
+    {
+      "epoch": 0.4747591522157996,
+      "grad_norm": 1.2628800873031072,
+      "learning_rate": 7.60493827160494e-05,
+      "loss": 0.6079,
+      "step": 154
+    },
+    {
+      "epoch": 0.47784200385356457,
+      "grad_norm": 3.4770621216815627,
+      "learning_rate": 7.654320987654322e-05,
+      "loss": 0.6347,
+      "step": 155
+    },
+    {
+      "epoch": 0.48092485549132946,
+      "grad_norm": 2.5486389481532443,
+      "learning_rate": 7.703703703703704e-05,
+      "loss": 0.639,
+      "step": 156
+    },
+    {
+      "epoch": 0.4840077071290944,
+      "grad_norm": 2.9976167232401396,
+      "learning_rate": 7.753086419753086e-05,
+      "loss": 0.6235,
+      "step": 157
+    },
+    {
+      "epoch": 0.48709055876685936,
+      "grad_norm": 2.3056233891635527,
+      "learning_rate": 7.80246913580247e-05,
+      "loss": 0.6216,
+      "step": 158
+    },
+    {
+      "epoch": 0.49017341040462425,
+      "grad_norm": 2.976038248788701,
+      "learning_rate": 7.851851851851852e-05,
+      "loss": 0.6251,
+      "step": 159
+    },
+    {
+      "epoch": 0.4932562620423892,
+      "grad_norm": 1.911291398877855,
+      "learning_rate": 7.901234567901235e-05,
+      "loss": 0.614,
+      "step": 160
+    },
+    {
+      "epoch": 0.49633911368015415,
+      "grad_norm": 2.649158593407182,
+      "learning_rate": 7.950617283950618e-05,
+      "loss": 0.6198,
+      "step": 161
+    },
+    {
+      "epoch": 0.49942196531791905,
+      "grad_norm": 1.9844788805877174,
+      "learning_rate": 8e-05,
+      "loss": 0.6197,
+      "step": 162
+    },
+    {
+      "epoch": 0.502504816955684,
+      "grad_norm": 2.6577295566999304,
+      "learning_rate": 7.999990714302639e-05,
+      "loss": 0.6158,
+      "step": 163
+    },
+    {
+      "epoch": 0.505587668593449,
+      "grad_norm": 2.0207525656317715,
+      "learning_rate": 7.999962857253669e-05,
+      "loss": 0.6174,
+      "step": 164
+    },
+    {
+      "epoch": 0.5086705202312138,
+      "grad_norm": 2.191900020660619,
+      "learning_rate": 7.999916428982425e-05,
+      "loss": 0.6132,
+      "step": 165
+    },
+    {
+      "epoch": 0.5117533718689788,
+      "grad_norm": 1.633631759032414,
+      "learning_rate": 7.999851429704464e-05,
+      "loss": 0.608,
+      "step": 166
+    },
+    {
+      "epoch": 0.5148362235067437,
+      "grad_norm": 2.42539803979078,
+      "learning_rate": 7.999767859721572e-05,
+      "loss": 0.6093,
+      "step": 167
+    },
+    {
+      "epoch": 0.5179190751445086,
+      "grad_norm": 1.837455748300822,
+      "learning_rate": 7.99966571942175e-05,
+      "loss": 0.6142,
+      "step": 168
+    },
+    {
+      "epoch": 0.5210019267822736,
+      "grad_norm": 2.077552727398867,
+      "learning_rate": 7.99954500927922e-05,
+      "loss": 0.6069,
+      "step": 169
+    },
+    {
+      "epoch": 0.5240847784200385,
+      "grad_norm": 1.7658840263859572,
+      "learning_rate": 7.999405729854421e-05,
+      "loss": 0.6016,
+      "step": 170
+    },
+    {
+      "epoch": 0.5271676300578034,
+      "grad_norm": 1.894419358367987,
+      "learning_rate": 7.999247881794007e-05,
+      "loss": 0.6007,
+      "step": 171
+    },
+    {
+      "epoch": 0.5302504816955684,
+      "grad_norm": 1.6208364650937428,
+      "learning_rate": 7.99907146583084e-05,
+      "loss": 0.605,
+      "step": 172
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 1.655132628381386,
+      "learning_rate": 7.998876482783996e-05,
+      "loss": 0.6045,
+      "step": 173
+    },
+    {
+      "epoch": 0.5364161849710982,
+      "grad_norm": 1.4961219657051947,
+      "learning_rate": 7.99866293355875e-05,
+      "loss": 0.5999,
+      "step": 174
+    },
+    {
+      "epoch": 0.5394990366088632,
+      "grad_norm": 1.5001935951646266,
+      "learning_rate": 7.99843081914658e-05,
+      "loss": 0.6001,
+      "step": 175
+    },
+    {
+      "epoch": 0.5425818882466281,
+      "grad_norm": 1.0242406158365422,
+      "learning_rate": 7.998180140625156e-05,
+      "loss": 0.5973,
+      "step": 176
+    },
+    {
+      "epoch": 0.545664739884393,
+      "grad_norm": 1.7556286880419985,
+      "learning_rate": 7.997910899158342e-05,
+      "loss": 0.6045,
+      "step": 177
+    },
+    {
+      "epoch": 0.548747591522158,
+      "grad_norm": 1.5918563665016492,
+      "learning_rate": 7.997623095996184e-05,
+      "loss": 0.5974,
+      "step": 178
+    },
+    {
+      "epoch": 0.5518304431599229,
+      "grad_norm": 1.8942858598284087,
+      "learning_rate": 7.997316732474911e-05,
+      "loss": 0.5912,
+      "step": 179
+    },
+    {
+      "epoch": 0.5549132947976878,
+      "grad_norm": 1.3136731253131897,
+      "learning_rate": 7.996991810016922e-05,
+      "loss": 0.5974,
+      "step": 180
+    },
+    {
+      "epoch": 0.5579961464354528,
+      "grad_norm": 1.4433781606780376,
+      "learning_rate": 7.996648330130779e-05,
+      "loss": 0.5951,
+      "step": 181
+    },
+    {
+      "epoch": 0.5610789980732177,
+      "grad_norm": 1.9139877061242987,
+      "learning_rate": 7.996286294411211e-05,
+      "loss": 0.6022,
+      "step": 182
+    },
+    {
+      "epoch": 0.5641618497109827,
+      "grad_norm": 1.8185903742736729,
+      "learning_rate": 7.995905704539095e-05,
+      "loss": 0.6044,
+      "step": 183
+    },
+    {
+      "epoch": 0.5672447013487476,
+      "grad_norm": 1.4921699392308978,
+      "learning_rate": 7.99550656228145e-05,
+      "loss": 0.6017,
+      "step": 184
+    },
+    {
+      "epoch": 0.5703275529865125,
+      "grad_norm": 1.319642412214861,
+      "learning_rate": 7.995088869491435e-05,
+      "loss": 0.5976,
+      "step": 185
+    },
+    {
+      "epoch": 0.5734104046242775,
+      "grad_norm": 1.3150278219430682,
+      "learning_rate": 7.994652628108334e-05,
+      "loss": 0.5962,
+      "step": 186
+    },
+    {
+      "epoch": 0.5764932562620424,
+      "grad_norm": 2.0403808045491383,
+      "learning_rate": 7.994197840157548e-05,
+      "loss": 0.6042,
+      "step": 187
+    },
+    {
+      "epoch": 0.5795761078998073,
+      "grad_norm": 0.9408463860648839,
+      "learning_rate": 7.993724507750591e-05,
+      "loss": 0.5965,
+      "step": 188
+    },
+    {
+      "epoch": 0.5826589595375723,
+      "grad_norm": 1.534350774197928,
+      "learning_rate": 7.993232633085074e-05,
+      "loss": 0.6167,
+      "step": 189
+    },
+    {
+      "epoch": 0.5857418111753372,
+      "grad_norm": 1.7239566096542989,
+      "learning_rate": 7.992722218444694e-05,
+      "loss": 0.6013,
+      "step": 190
+    },
+    {
+      "epoch": 0.5888246628131021,
+      "grad_norm": 1.195728189473519,
+      "learning_rate": 7.992193266199231e-05,
+      "loss": 0.6004,
+      "step": 191
+    },
+    {
+      "epoch": 0.5919075144508671,
+      "grad_norm": 1.6217442570229952,
+      "learning_rate": 7.991645778804529e-05,
+      "loss": 0.6132,
+      "step": 192
+    },
+    {
+      "epoch": 0.594990366088632,
+      "grad_norm": 1.1074294077623117,
+      "learning_rate": 7.99107975880249e-05,
+      "loss": 0.6026,
+      "step": 193
+    },
+    {
+      "epoch": 0.5980732177263969,
+      "grad_norm": 1.631691774632524,
+      "learning_rate": 7.99049520882106e-05,
+      "loss": 0.603,
+      "step": 194
+    },
+    {
+      "epoch": 0.6011560693641619,
+      "grad_norm": 1.5147111634612662,
+      "learning_rate": 7.989892131574214e-05,
+      "loss": 0.6045,
+      "step": 195
+    },
+    {
+      "epoch": 0.6042389210019268,
+      "grad_norm": 1.05344099155785,
+      "learning_rate": 7.989270529861949e-05,
+      "loss": 0.5993,
+      "step": 196
+    },
+    {
+      "epoch": 0.6073217726396917,
+      "grad_norm": 1.8754914330123993,
+      "learning_rate": 7.98863040657027e-05,
+      "loss": 0.5979,
+      "step": 197
+    },
+    {
+      "epoch": 0.6104046242774567,
+      "grad_norm": 1.147440527781351,
+      "learning_rate": 7.987971764671168e-05,
+      "loss": 0.6005,
+      "step": 198
+    },
+    {
+      "epoch": 0.6134874759152216,
+      "grad_norm": 1.827350384720111,
+      "learning_rate": 7.987294607222624e-05,
+      "loss": 0.5967,
+      "step": 199
+    },
+    {
+      "epoch": 0.6165703275529865,
+      "grad_norm": 1.4984285077077433,
+      "learning_rate": 7.986598937368572e-05,
+      "loss": 0.6054,
+      "step": 200
+    },
+    {
+      "epoch": 0.6196531791907515,
+      "grad_norm": 1.0711487251789733,
+      "learning_rate": 7.985884758338903e-05,
+      "loss": 0.5972,
+      "step": 201
+    },
+    {
+      "epoch": 0.6227360308285164,
+      "grad_norm": 1.0551049774791001,
+      "learning_rate": 7.985152073449444e-05,
+      "loss": 0.5943,
+      "step": 202
+    },
+    {
+      "epoch": 0.6258188824662813,
+      "grad_norm": 1.2848273360983933,
+      "learning_rate": 7.98440088610194e-05,
+      "loss": 0.5994,
+      "step": 203
+    },
+    {
+      "epoch": 0.6289017341040463,
+      "grad_norm": 1.6224456877226094,
+      "learning_rate": 7.983631199784038e-05,
+      "loss": 0.5945,
+      "step": 204
+    },
+    {
+      "epoch": 0.6319845857418112,
+      "grad_norm": 1.2413684466471406,
+      "learning_rate": 7.982843018069277e-05,
+      "loss": 0.5901,
+      "step": 205
+    },
+    {
+      "epoch": 0.6350674373795762,
+      "grad_norm": 1.5377565658885557,
+      "learning_rate": 7.982036344617065e-05,
+      "loss": 0.5909,
+      "step": 206
+    },
+    {
+      "epoch": 0.638150289017341,
+      "grad_norm": 1.3421927009040955,
+      "learning_rate": 7.981211183172663e-05,
+      "loss": 0.589,
+      "step": 207
+    },
+    {
+      "epoch": 0.641233140655106,
+      "grad_norm": 1.0738882180184097,
+      "learning_rate": 7.980367537567175e-05,
+      "loss": 0.588,
+      "step": 208
+    },
+    {
+      "epoch": 0.644315992292871,
+      "grad_norm": 1.2682976978702023,
+      "learning_rate": 7.979505411717516e-05,
+      "loss": 0.5895,
+      "step": 209
+    },
+    {
+      "epoch": 0.6473988439306358,
+      "grad_norm": 1.329406883612844,
+      "learning_rate": 7.978624809626407e-05,
+      "loss": 0.5906,
+      "step": 210
+    },
+    {
+      "epoch": 0.6504816955684007,
+      "grad_norm": 1.5266402187568815,
+      "learning_rate": 7.977725735382349e-05,
+      "loss": 0.5866,
+      "step": 211
+    },
+    {
+      "epoch": 0.6535645472061657,
+      "grad_norm": 0.7488118706101616,
+      "learning_rate": 7.976808193159611e-05,
+      "loss": 0.5868,
+      "step": 212
+    },
+    {
+      "epoch": 0.6566473988439306,
+      "grad_norm": 1.4352626914409072,
+      "learning_rate": 7.9758721872182e-05,
+      "loss": 0.5818,
+      "step": 213
+    },
+    {
+      "epoch": 0.6597302504816955,
+      "grad_norm": 1.4119414599194002,
+      "learning_rate": 7.97491772190385e-05,
+      "loss": 0.5868,
+      "step": 214
+    },
+    {
+      "epoch": 0.6628131021194605,
+      "grad_norm": 0.6937099084342753,
+      "learning_rate": 7.973944801648001e-05,
+      "loss": 0.5816,
+      "step": 215
+    },
+    {
+      "epoch": 0.6658959537572254,
+      "grad_norm": 1.139611437599005,
+      "learning_rate": 7.972953430967773e-05,
+      "loss": 0.5855,
+      "step": 216
+    },
+    {
+      "epoch": 0.6689788053949903,
+      "grad_norm": 1.7588598887369369,
+      "learning_rate": 7.971943614465949e-05,
+      "loss": 0.5956,
+      "step": 217
+    },
+    {
+      "epoch": 0.6720616570327553,
+      "grad_norm": 1.3804947001854182,
+      "learning_rate": 7.970915356830956e-05,
+      "loss": 0.582,
+      "step": 218
+    },
+    {
+      "epoch": 0.6751445086705202,
+      "grad_norm": 1.5546158957849832,
+      "learning_rate": 7.969868662836839e-05,
+      "loss": 0.5825,
+      "step": 219
+    },
+    {
+      "epoch": 0.6782273603082851,
+      "grad_norm": 0.991254051505734,
+      "learning_rate": 7.968803537343239e-05,
+      "loss": 0.5843,
+      "step": 220
+    },
+    {
+      "epoch": 0.6813102119460501,
+      "grad_norm": 2.7906603027998793,
+      "learning_rate": 7.967719985295372e-05,
+      "loss": 0.5999,
+      "step": 221
+    },
+    {
+      "epoch": 0.684393063583815,
+      "grad_norm": 1.7848553097909525,
+      "learning_rate": 7.966618011724004e-05,
+      "loss": 0.5968,
+      "step": 222
+    },
+    {
+      "epoch": 0.6874759152215799,
+      "grad_norm": 2.367140342144655,
+      "learning_rate": 7.965497621745437e-05,
+      "loss": 0.5916,
+      "step": 223
+    },
+    {
+      "epoch": 0.6905587668593449,
+      "grad_norm": 1.8291521301648463,
+      "learning_rate": 7.964358820561469e-05,
+      "loss": 0.5996,
+      "step": 224
+    },
+    {
+      "epoch": 0.6936416184971098,
+      "grad_norm": 1.7931577769369804,
+      "learning_rate": 7.963201613459381e-05,
+      "loss": 0.5916,
+      "step": 225
+    },
+    {
+      "epoch": 0.6967244701348747,
+      "grad_norm": 0.9889883119566096,
+      "learning_rate": 7.962026005811914e-05,
+      "loss": 0.5854,
+      "step": 226
+    },
+    {
+      "epoch": 0.6998073217726397,
+      "grad_norm": 1.5236346644059422,
+      "learning_rate": 7.96083200307723e-05,
+      "loss": 0.5921,
+      "step": 227
+    },
+    {
+      "epoch": 0.7028901734104046,
+      "grad_norm": 1.0993899360354793,
+      "learning_rate": 7.959619610798909e-05,
+      "loss": 0.5874,
+      "step": 228
+    },
+    {
+      "epoch": 0.7059730250481696,
+      "grad_norm": 1.3820399982937879,
+      "learning_rate": 7.958388834605901e-05,
+      "loss": 0.5826,
+      "step": 229
+    },
+    {
+      "epoch": 0.7090558766859345,
+      "grad_norm": 1.1656132386771638,
+      "learning_rate": 7.957139680212518e-05,
+      "loss": 0.5865,
+      "step": 230
+    },
+    {
+      "epoch": 0.7121387283236994,
+      "grad_norm": 0.9027958773244722,
+      "learning_rate": 7.95587215341839e-05,
+      "loss": 0.5779,
+      "step": 231
+    },
+    {
+      "epoch": 0.7152215799614644,
+      "grad_norm": 1.3064982091334527,
+      "learning_rate": 7.954586260108454e-05,
+      "loss": 0.5852,
+      "step": 232
+    },
+    {
+      "epoch": 0.7183044315992293,
+      "grad_norm": 0.824109578479739,
+      "learning_rate": 7.953282006252919e-05,
+      "loss": 0.5825,
+      "step": 233
+    },
+    {
+      "epoch": 0.7213872832369942,
+      "grad_norm": 0.8979704816960984,
+      "learning_rate": 7.951959397907237e-05,
+      "loss": 0.5812,
+      "step": 234
+    },
+    {
+      "epoch": 0.7244701348747592,
+      "grad_norm": 1.3137609948846414,
+      "learning_rate": 7.950618441212078e-05,
+      "loss": 0.5838,
+      "step": 235
+    },
+    {
+      "epoch": 0.7275529865125241,
+      "grad_norm": 1.4947719300203344,
+      "learning_rate": 7.949259142393305e-05,
+      "loss": 0.583,
+      "step": 236
+    },
+    {
+      "epoch": 0.730635838150289,
+      "grad_norm": 1.0993997214832212,
+      "learning_rate": 7.947881507761932e-05,
+      "loss": 0.5817,
+      "step": 237
+    },
+    {
+      "epoch": 0.733718689788054,
+      "grad_norm": 0.954803719816968,
+      "learning_rate": 7.946485543714108e-05,
+      "loss": 0.5801,
+      "step": 238
+    },
+    {
+      "epoch": 0.7368015414258189,
+      "grad_norm": 1.2754064634359743,
+      "learning_rate": 7.945071256731087e-05,
+      "loss": 0.5796,
+      "step": 239
+    },
+    {
+      "epoch": 0.7398843930635838,
+      "grad_norm": 0.9770128775618113,
+      "learning_rate": 7.943638653379186e-05,
+      "loss": 0.5747,
+      "step": 240
+    },
+    {
+      "epoch": 0.7429672447013488,
+      "grad_norm": 1.5391615029555052,
+      "learning_rate": 7.942187740309768e-05,
+      "loss": 0.5795,
+      "step": 241
+    },
+    {
+      "epoch": 0.7460500963391137,
+      "grad_norm": 1.2347561842025963,
+      "learning_rate": 7.940718524259201e-05,
+      "loss": 0.5773,
+      "step": 242
+    },
+    {
+      "epoch": 0.7491329479768786,
+      "grad_norm": 0.9139339011556599,
+      "learning_rate": 7.939231012048833e-05,
+      "loss": 0.5736,
+      "step": 243
+    },
+    {
+      "epoch": 0.7522157996146436,
+      "grad_norm": 1.7386157834395892,
+      "learning_rate": 7.937725210584958e-05,
+      "loss": 0.5845,
+      "step": 244
+    },
+    {
+      "epoch": 0.7552986512524085,
+      "grad_norm": 0.9528511653131708,
+      "learning_rate": 7.936201126858786e-05,
+      "loss": 0.5776,
+      "step": 245
+    },
+    {
+      "epoch": 0.7583815028901734,
+      "grad_norm": 1.5937482368903735,
+      "learning_rate": 7.934658767946405e-05,
+      "loss": 0.5803,
+      "step": 246
+    },
+    {
+      "epoch": 0.7614643545279384,
+      "grad_norm": 0.9157855007725487,
+      "learning_rate": 7.933098141008754e-05,
+      "loss": 0.5784,
+      "step": 247
+    },
+    {
+      "epoch": 0.7645472061657033,
+      "grad_norm": 1.796526188766546,
+      "learning_rate": 7.93151925329159e-05,
+      "loss": 0.5837,
+      "step": 248
+    },
+    {
+      "epoch": 0.7676300578034682,
+      "grad_norm": 1.067545515831627,
+      "learning_rate": 7.92992211212545e-05,
+      "loss": 0.5825,
+      "step": 249
+    },
+    {
+      "epoch": 0.7707129094412332,
+      "grad_norm": 1.75243679239792,
+      "learning_rate": 7.928306724925615e-05,
+      "loss": 0.5872,
+      "step": 250
+    },
+    {
+      "epoch": 0.7737957610789981,
+      "grad_norm": 1.246580955992929,
+      "learning_rate": 7.926673099192085e-05,
+      "loss": 0.5899,
+      "step": 251
+    },
+    {
+      "epoch": 0.776878612716763,
+      "grad_norm": 1.4842464216305902,
+      "learning_rate": 7.925021242509539e-05,
+      "loss": 0.5827,
+      "step": 252
+    },
+    {
+      "epoch": 0.779961464354528,
+      "grad_norm": 1.1379361219065842,
+      "learning_rate": 7.923351162547295e-05,
+      "loss": 0.5861,
+      "step": 253
+    },
+    {
+      "epoch": 0.7830443159922928,
+      "grad_norm": 1.0238695774536577,
+      "learning_rate": 7.921662867059285e-05,
+      "loss": 0.5738,
+      "step": 254
+    },
+    {
+      "epoch": 0.7861271676300579,
+      "grad_norm": 1.113201949538522,
+      "learning_rate": 7.919956363884005e-05,
+      "loss": 0.5788,
+      "step": 255
+    },
+    {
+      "epoch": 0.7892100192678227,
+      "grad_norm": 0.8382973915827036,
+      "learning_rate": 7.918231660944493e-05,
+      "loss": 0.5809,
+      "step": 256
+    },
+    {
+      "epoch": 0.7922928709055876,
+      "grad_norm": 0.8271683342239057,
+      "learning_rate": 7.916488766248285e-05,
+      "loss": 0.5785,
+      "step": 257
+    },
+    {
+      "epoch": 0.7953757225433526,
+      "grad_norm": 0.9982661342228444,
+      "learning_rate": 7.914727687887377e-05,
+      "loss": 0.5748,
+      "step": 258
+    },
+    {
+      "epoch": 0.7984585741811175,
+      "grad_norm": 1.4454961867701526,
+      "learning_rate": 7.912948434038188e-05,
+      "loss": 0.5761,
+      "step": 259
+    },
+    {
+      "epoch": 0.8015414258188824,
+      "grad_norm": 1.0220822168108143,
+      "learning_rate": 7.911151012961526e-05,
+      "loss": 0.577,
+      "step": 260
+    },
+    {
+      "epoch": 0.8046242774566474,
+      "grad_norm": 1.258740036514348,
+      "learning_rate": 7.909335433002543e-05,
+      "loss": 0.5751,
+      "step": 261
+    },
+    {
+      "epoch": 0.8077071290944123,
+      "grad_norm": 1.3258682043764047,
+      "learning_rate": 7.907501702590702e-05,
+      "loss": 0.5775,
+      "step": 262
+    },
+    {
+      "epoch": 0.8107899807321772,
+      "grad_norm": 0.8008704431828687,
+      "learning_rate": 7.905649830239739e-05,
+      "loss": 0.5715,
+      "step": 263
+    },
+    {
+      "epoch": 0.8138728323699422,
+      "grad_norm": 0.9108543766937637,
+      "learning_rate": 7.903779824547614e-05,
+      "loss": 0.5735,
+      "step": 264
+    },
+    {
+      "epoch": 0.8169556840077071,
+      "grad_norm": 1.324370884784008,
+      "learning_rate": 7.901891694196483e-05,
+      "loss": 0.579,
+      "step": 265
+    },
+    {
+      "epoch": 0.820038535645472,
+      "grad_norm": 1.1194610334743726,
+      "learning_rate": 7.899985447952648e-05,
+      "loss": 0.5729,
+      "step": 266
+    },
+    {
+      "epoch": 0.823121387283237,
+      "grad_norm": 1.0788872436660464,
+      "learning_rate": 7.898061094666521e-05,
+      "loss": 0.5753,
+      "step": 267
+    },
+    {
+      "epoch": 0.8262042389210019,
+      "grad_norm": 1.75060108883478,
+      "learning_rate": 7.896118643272584e-05,
+      "loss": 0.5785,
+      "step": 268
+    },
+    {
+      "epoch": 0.8292870905587668,
+      "grad_norm": 0.6706029407953169,
+      "learning_rate": 7.894158102789344e-05,
+      "loss": 0.5714,
+      "step": 269
+    },
+    {
+      "epoch": 0.8323699421965318,
+      "grad_norm": 2.059014468269478,
+      "learning_rate": 7.892179482319297e-05,
+      "loss": 0.5842,
+      "step": 270
+    },
+    {
+      "epoch": 0.8354527938342967,
+      "grad_norm": 1.1260373150835865,
+      "learning_rate": 7.890182791048875e-05,
+      "loss": 0.5782,
+      "step": 271
+    },
+    {
+      "epoch": 0.8385356454720616,
+      "grad_norm": 1.8831605520021648,
+      "learning_rate": 7.888168038248414e-05,
+      "loss": 0.5907,
+      "step": 272
+    },
+    {
+      "epoch": 0.8416184971098266,
+      "grad_norm": 1.5909106614322919,
+      "learning_rate": 7.886135233272108e-05,
+      "loss": 0.5855,
+      "step": 273
+    },
+    {
+      "epoch": 0.8447013487475915,
+      "grad_norm": 1.5620850035892349,
+      "learning_rate": 7.88408438555796e-05,
+      "loss": 0.5804,
+      "step": 274
+    },
+    {
+      "epoch": 0.8477842003853564,
+      "grad_norm": 1.5009206737228515,
+      "learning_rate": 7.882015504627749e-05,
+      "loss": 0.5835,
+      "step": 275
+    },
+    {
+      "epoch": 0.8508670520231214,
+      "grad_norm": 1.371211854984156,
+      "learning_rate": 7.879928600086974e-05,
+      "loss": 0.5725,
+      "step": 276
+    },
+    {
+      "epoch": 0.8539499036608863,
+      "grad_norm": 1.5464920439804397,
+      "learning_rate": 7.87782368162482e-05,
+      "loss": 0.5735,
+      "step": 277
+    },
+    {
+      "epoch": 0.8570327552986513,
+      "grad_norm": 1.288380001368499,
+      "learning_rate": 7.875700759014099e-05,
+      "loss": 0.5748,
+      "step": 278
+    },
+    {
+      "epoch": 0.8601156069364162,
+      "grad_norm": 1.369018500620383,
+      "learning_rate": 7.873559842111225e-05,
+      "loss": 0.5754,
+      "step": 279
+    },
+    {
+      "epoch": 0.8631984585741811,
+      "grad_norm": 0.9860978537471357,
+      "learning_rate": 7.871400940856147e-05,
+      "loss": 0.5697,
+      "step": 280
+    },
+    {
+      "epoch": 0.8662813102119461,
+      "grad_norm": 1.435898378737718,
+      "learning_rate": 7.869224065272321e-05,
+      "loss": 0.5677,
+      "step": 281
+    },
+    {
+      "epoch": 0.869364161849711,
+      "grad_norm": 0.8935369915013904,
+      "learning_rate": 7.867029225466649e-05,
+      "loss": 0.5725,
+      "step": 282
+    },
+    {
+      "epoch": 0.8724470134874759,
+      "grad_norm": 1.2749107745374528,
+      "learning_rate": 7.864816431629439e-05,
+      "loss": 0.5664,
+      "step": 283
+    },
+    {
+      "epoch": 0.8755298651252409,
+      "grad_norm": 0.7998662041187654,
+      "learning_rate": 7.862585694034358e-05,
+      "loss": 0.5693,
+      "step": 284
+    },
+    {
+      "epoch": 0.8786127167630058,
+      "grad_norm": 0.7778458486807098,
+      "learning_rate": 7.860337023038387e-05,
+      "loss": 0.5681,
+      "step": 285
+    },
+    {
+      "epoch": 0.8816955684007707,
+      "grad_norm": 1.0439313827705001,
+      "learning_rate": 7.858070429081761e-05,
+      "loss": 0.5701,
+      "step": 286
+    },
+    {
+      "epoch": 0.8847784200385357,
+      "grad_norm": 0.9068847676907831,
+      "learning_rate": 7.855785922687933e-05,
+      "loss": 0.561,
+      "step": 287
+    },
+    {
+      "epoch": 0.8878612716763006,
+      "grad_norm": 0.8400874254217158,
+      "learning_rate": 7.853483514463521e-05,
+      "loss": 0.5652,
+      "step": 288
+    },
+    {
+      "epoch": 0.8909441233140655,
+      "grad_norm": 0.9891521129841898,
+      "learning_rate": 7.851163215098261e-05,
+      "loss": 0.564,
+      "step": 289
+    },
+    {
+      "epoch": 0.8940269749518305,
+      "grad_norm": 1.7320906830329201,
+      "learning_rate": 7.848825035364947e-05,
+      "loss": 0.5733,
+      "step": 290
+    },
+    {
+      "epoch": 0.8971098265895954,
+      "grad_norm": 0.769250448302101,
+      "learning_rate": 7.846468986119397e-05,
+      "loss": 0.5647,
+      "step": 291
+    },
+    {
+      "epoch": 0.9001926782273603,
+      "grad_norm": 1.1961802382699547,
+      "learning_rate": 7.84409507830039e-05,
+      "loss": 0.5667,
+      "step": 292
+    },
+    {
+      "epoch": 0.9032755298651253,
+      "grad_norm": 1.5391485868303119,
+      "learning_rate": 7.841703322929622e-05,
+      "loss": 0.5704,
+      "step": 293
+    },
+    {
+      "epoch": 0.9063583815028902,
+      "grad_norm": 0.7774096039145763,
+      "learning_rate": 7.839293731111648e-05,
+      "loss": 0.564,
+      "step": 294
+    },
+    {
+      "epoch": 0.9094412331406551,
+      "grad_norm": 1.3750399761963137,
+      "learning_rate": 7.836866314033842e-05,
+      "loss": 0.5727,
+      "step": 295
+    },
+    {
+      "epoch": 0.9125240847784201,
+      "grad_norm": 0.8951291009718272,
+      "learning_rate": 7.834421082966332e-05,
+      "loss": 0.5669,
+      "step": 296
+    },
+    {
+      "epoch": 0.915606936416185,
+      "grad_norm": 1.0077135390736174,
+      "learning_rate": 7.831958049261956e-05,
+      "loss": 0.5664,
+      "step": 297
+    },
+    {
+      "epoch": 0.9186897880539499,
+      "grad_norm": 1.2034659479055732,
+      "learning_rate": 7.829477224356207e-05,
+      "loss": 0.5692,
+      "step": 298
+    },
+    {
+      "epoch": 0.9217726396917149,
+      "grad_norm": 0.8550199343779012,
+      "learning_rate": 7.82697861976718e-05,
+      "loss": 0.5617,
+      "step": 299
+    },
+    {
+      "epoch": 0.9248554913294798,
+      "grad_norm": 0.8274911336207205,
+      "learning_rate": 7.82446224709552e-05,
+      "loss": 0.5665,
+      "step": 300
+    },
+    {
+      "epoch": 0.9279383429672448,
+      "grad_norm": 0.8633320367878546,
+      "learning_rate": 7.82192811802436e-05,
+      "loss": 0.5637,
+      "step": 301
+    },
+    {
+      "epoch": 0.9310211946050096,
+      "grad_norm": 1.2472620909139571,
+      "learning_rate": 7.81937624431928e-05,
+      "loss": 0.564,
+      "step": 302
+    },
+    {
+      "epoch": 0.9341040462427745,
+      "grad_norm": 0.7793254345235272,
+      "learning_rate": 7.816806637828246e-05,
+      "loss": 0.5626,
+      "step": 303
+    },
+    {
+      "epoch": 0.9371868978805395,
+      "grad_norm": 0.8510822717833507,
+      "learning_rate": 7.814219310481547e-05,
+      "loss": 0.5581,
+      "step": 304
+    },
+    {
+      "epoch": 0.9402697495183044,
+      "grad_norm": 0.9440471390979412,
+      "learning_rate": 7.811614274291757e-05,
+      "loss": 0.5604,
+      "step": 305
+    },
+    {
+      "epoch": 0.9433526011560693,
+      "grad_norm": 0.9853741760948094,
+      "learning_rate": 7.808991541353662e-05,
+      "loss": 0.5589,
+      "step": 306
+    },
+    {
+      "epoch": 0.9464354527938343,
+      "grad_norm": 1.1375385668546794,
+      "learning_rate": 7.806351123844215e-05,
+      "loss": 0.5637,
+      "step": 307
+    },
+    {
+      "epoch": 0.9495183044315992,
+      "grad_norm": 1.0721316364338647,
+      "learning_rate": 7.803693034022475e-05,
+      "loss": 0.5607,
+      "step": 308
+    },
+    {
+      "epoch": 0.9526011560693641,
+      "grad_norm": 1.2668934466340183,
+      "learning_rate": 7.801017284229552e-05,
+      "loss": 0.559,
+      "step": 309
+    },
+    {
+      "epoch": 0.9556840077071291,
+      "grad_norm": 0.5522081277664963,
+      "learning_rate": 7.798323886888545e-05,
+      "loss": 0.5559,
+      "step": 310
+    },
+    {
+      "epoch": 0.958766859344894,
+      "grad_norm": 0.7033492787648649,
+      "learning_rate": 7.795612854504493e-05,
+      "loss": 0.5569,
+      "step": 311
+    },
+    {
+      "epoch": 0.9618497109826589,
+      "grad_norm": 0.8494344394600313,
+      "learning_rate": 7.792884199664307e-05,
+      "loss": 0.5574,
+      "step": 312
+    },
+    {
+      "epoch": 0.9649325626204239,
+      "grad_norm": 1.2152570299023822,
+      "learning_rate": 7.790137935036718e-05,
+      "loss": 0.5621,
+      "step": 313
+    },
+    {
+      "epoch": 0.9680154142581888,
+      "grad_norm": 0.8998789658239084,
+      "learning_rate": 7.787374073372221e-05,
+      "loss": 0.5544,
+      "step": 314
+    },
+    {
+      "epoch": 0.9710982658959537,
+      "grad_norm": 0.8608777407027878,
+      "learning_rate": 7.784592627503004e-05,
+      "loss": 0.5562,
+      "step": 315
+    },
+    {
+      "epoch": 0.9741811175337187,
+      "grad_norm": 0.9213970840028392,
+      "learning_rate": 7.781793610342897e-05,
+      "loss": 0.5564,
+      "step": 316
+    },
+    {
+      "epoch": 0.9772639691714836,
+      "grad_norm": 0.9698381497921442,
+      "learning_rate": 7.77897703488732e-05,
+      "loss": 0.5571,
+      "step": 317
+    },
+    {
+      "epoch": 0.9803468208092485,
+      "grad_norm": 1.2344314308265052,
+      "learning_rate": 7.7761429142132e-05,
+      "loss": 0.5625,
+      "step": 318
+    },
+    {
+      "epoch": 0.9834296724470135,
+      "grad_norm": 0.8454154467150172,
+      "learning_rate": 7.773291261478936e-05,
+      "loss": 0.5589,
+      "step": 319
+    },
+    {
+      "epoch": 0.9865125240847784,
+      "grad_norm": 0.818914610328381,
+      "learning_rate": 7.770422089924315e-05,
+      "loss": 0.5599,
+      "step": 320
+    },
+    {
+      "epoch": 0.9895953757225433,
+      "grad_norm": 0.8746933498512478,
+      "learning_rate": 7.767535412870469e-05,
+      "loss": 0.5617,
+      "step": 321
+    },
+    {
+      "epoch": 0.9926782273603083,
+      "grad_norm": 0.9347296430477423,
+      "learning_rate": 7.764631243719804e-05,
+      "loss": 0.5599,
+      "step": 322
+    },
+    {
+      "epoch": 0.9957610789980732,
+      "grad_norm": 0.9198100961787572,
+      "learning_rate": 7.761709595955936e-05,
+      "loss": 0.5575,
+      "step": 323
+    },
+    {
+      "epoch": 0.9988439306358381,
+      "grad_norm": 1.0449248532647026,
+      "learning_rate": 7.758770483143634e-05,
+      "loss": 0.5599,
+      "step": 324
+    },
+    {
+      "epoch": 1.001926782273603,
+      "grad_norm": 2.1348799419427413,
+      "learning_rate": 7.755813918928754e-05,
+      "loss": 0.9006,
+      "step": 325
+    },
+    {
+      "epoch": 1.005009633911368,
+      "grad_norm": 0.7781225544637964,
+      "learning_rate": 7.752839917038176e-05,
+      "loss": 0.5409,
+      "step": 326
+    },
+    {
+      "epoch": 1.0080924855491329,
+      "grad_norm": 0.7247163718442423,
+      "learning_rate": 7.749848491279741e-05,
+      "loss": 0.541,
+      "step": 327
+    },
+    {
+      "epoch": 1.011175337186898,
+      "grad_norm": 0.8674239793358375,
+      "learning_rate": 7.746839655542187e-05,
+      "loss": 0.5437,
+      "step": 328
+    },
+    {
+      "epoch": 1.014258188824663,
+      "grad_norm": 1.2851053917966024,
+      "learning_rate": 7.74381342379508e-05,
+      "loss": 0.5467,
+      "step": 329
+    },
+    {
+      "epoch": 1.0173410404624277,
+      "grad_norm": 0.7922998349676916,
+      "learning_rate": 7.74076981008876e-05,
+      "loss": 0.5428,
+      "step": 330
+    },
+    {
+      "epoch": 1.0204238921001927,
+      "grad_norm": 0.7435375211325687,
+      "learning_rate": 7.737708828554261e-05,
+      "loss": 0.5397,
+      "step": 331
+    },
+    {
+      "epoch": 1.0235067437379577,
+      "grad_norm": 0.970964978035143,
+      "learning_rate": 7.734630493403262e-05,
+      "loss": 0.5418,
+      "step": 332
+    },
+    {
+      "epoch": 1.0265895953757225,
+      "grad_norm": 0.9289911168548828,
+      "learning_rate": 7.731534818928004e-05,
+      "loss": 0.5437,
+      "step": 333
+    },
+    {
+      "epoch": 1.0296724470134875,
+      "grad_norm": 0.8408605923106182,
+      "learning_rate": 7.728421819501232e-05,
+      "loss": 0.5462,
+      "step": 334
+    },
+    {
+      "epoch": 1.0327552986512525,
+      "grad_norm": 0.7314104111298062,
+      "learning_rate": 7.725291509576139e-05,
+      "loss": 0.5417,
+      "step": 335
+    },
+    {
+      "epoch": 1.0358381502890173,
+      "grad_norm": 0.5543890744417052,
+      "learning_rate": 7.722143903686274e-05,
+      "loss": 0.5372,
+      "step": 336
+    },
+    {
+      "epoch": 1.0389210019267823,
+      "grad_norm": 0.6727866939033414,
+      "learning_rate": 7.7189790164455e-05,
+      "loss": 0.5389,
+      "step": 337
+    },
+    {
+      "epoch": 1.0420038535645473,
+      "grad_norm": 0.7241288134455114,
+      "learning_rate": 7.715796862547903e-05,
+      "loss": 0.5426,
+      "step": 338
+    },
+    {
+      "epoch": 1.045086705202312,
+      "grad_norm": 0.6934541025509483,
+      "learning_rate": 7.712597456767747e-05,
+      "loss": 0.5339,
+      "step": 339
+    },
+    {
+      "epoch": 1.048169556840077,
+      "grad_norm": 0.693001061172053,
+      "learning_rate": 7.709380813959389e-05,
+      "loss": 0.5449,
+      "step": 340
+    },
+    {
+      "epoch": 1.051252408477842,
+      "grad_norm": 0.9573876629636278,
+      "learning_rate": 7.706146949057211e-05,
+      "loss": 0.5343,
+      "step": 341
+    },
+    {
+      "epoch": 1.0543352601156069,
+      "grad_norm": 1.2740258417892243,
+      "learning_rate": 7.702895877075563e-05,
+      "loss": 0.5389,
+      "step": 342
+    },
+    {
+      "epoch": 1.0574181117533719,
+      "grad_norm": 0.8780933952057481,
+      "learning_rate": 7.699627613108677e-05,
+      "loss": 0.5393,
+      "step": 343
+    },
+    {
+      "epoch": 1.0605009633911369,
+      "grad_norm": 0.9117155028935067,
+      "learning_rate": 7.696342172330607e-05,
+      "loss": 0.5358,
+      "step": 344
+    },
+    {
+      "epoch": 1.0635838150289016,
+      "grad_norm": 0.9180036803559745,
+      "learning_rate": 7.693039569995162e-05,
+      "loss": 0.5415,
+      "step": 345
+    },
+    {
+      "epoch": 1.0666666666666667,
+      "grad_norm": 1.3123668505963522,
+      "learning_rate": 7.689719821435821e-05,
+      "loss": 0.5396,
+      "step": 346
+    },
+    {
+      "epoch": 1.0697495183044317,
+      "grad_norm": 0.4645379250221101,
+      "learning_rate": 7.686382942065678e-05,
+      "loss": 0.5386,
+      "step": 347
+    },
+    {
+      "epoch": 1.0728323699421964,
+      "grad_norm": 0.6446688050436276,
+      "learning_rate": 7.683028947377355e-05,
+      "loss": 0.5432,
+      "step": 348
+    },
+    {
+      "epoch": 1.0759152215799614,
+      "grad_norm": 0.8937545976524452,
+      "learning_rate": 7.679657852942943e-05,
+      "loss": 0.5381,
+      "step": 349
+    },
+    {
+      "epoch": 1.0789980732177264,
+      "grad_norm": 1.1765605963859298,
+      "learning_rate": 7.676269674413923e-05,
+      "loss": 0.5384,
+      "step": 350
+    },
+    {
+      "epoch": 1.0820809248554912,
+      "grad_norm": 0.9563821514263772,
+      "learning_rate": 7.672864427521097e-05,
+      "loss": 0.5377,
+      "step": 351
+    },
+    {
+      "epoch": 1.0851637764932562,
+      "grad_norm": 0.8690655728928455,
+      "learning_rate": 7.66944212807451e-05,
+      "loss": 0.5395,
+      "step": 352
+    },
+    {
+      "epoch": 1.0882466281310212,
+      "grad_norm": 0.7951284905846605,
+      "learning_rate": 7.66600279196338e-05,
+      "loss": 0.5339,
+      "step": 353
+    },
+    {
+      "epoch": 1.091329479768786,
+      "grad_norm": 0.9576230781522321,
+      "learning_rate": 7.662546435156025e-05,
+      "loss": 0.5381,
+      "step": 354
+    },
+    {
+      "epoch": 1.094412331406551,
+      "grad_norm": 1.4115239538918096,
+      "learning_rate": 7.659073073699785e-05,
+      "loss": 0.5352,
+      "step": 355
+    },
+    {
+      "epoch": 1.097495183044316,
+      "grad_norm": 0.5970597530516887,
+      "learning_rate": 7.655582723720953e-05,
+      "loss": 0.5419,
+      "step": 356
+    },
+    {
+      "epoch": 1.1005780346820808,
+      "grad_norm": 0.8553145770732488,
+      "learning_rate": 7.652075401424697e-05,
+      "loss": 0.5363,
+      "step": 357
+    },
+    {
+      "epoch": 1.1036608863198458,
+      "grad_norm": 1.2865734454031699,
+      "learning_rate": 7.648551123094981e-05,
+      "loss": 0.541,
+      "step": 358
+    },
+    {
+      "epoch": 1.1067437379576108,
+      "grad_norm": 0.7775758076077842,
+      "learning_rate": 7.645009905094499e-05,
+      "loss": 0.5346,
+      "step": 359
+    },
+    {
+      "epoch": 1.1098265895953756,
+      "grad_norm": 0.8828002009012496,
+      "learning_rate": 7.641451763864587e-05,
+      "loss": 0.536,
+      "step": 360
+    },
+    {
+      "epoch": 1.1129094412331406,
+      "grad_norm": 0.9000648684754602,
+      "learning_rate": 7.637876715925158e-05,
+      "loss": 0.5354,
+      "step": 361
+    },
+    {
+      "epoch": 1.1159922928709056,
+      "grad_norm": 0.6532917321316766,
+      "learning_rate": 7.63428477787462e-05,
+      "loss": 0.5405,
+      "step": 362
+    },
+    {
+      "epoch": 1.1190751445086704,
+      "grad_norm": 0.8233928580557633,
+      "learning_rate": 7.630675966389796e-05,
+      "loss": 0.5356,
+      "step": 363
+    },
+    {
+      "epoch": 1.1221579961464354,
+      "grad_norm": 1.286204152597659,
+      "learning_rate": 7.627050298225852e-05,
+      "loss": 0.5356,
+      "step": 364
+    },
+    {
+      "epoch": 1.1252408477842004,
+      "grad_norm": 0.671031297932369,
+      "learning_rate": 7.623407790216215e-05,
+      "loss": 0.5374,
+      "step": 365
+    },
+    {
+      "epoch": 1.1283236994219652,
+      "grad_norm": 0.6073254592895291,
+      "learning_rate": 7.619748459272501e-05,
+      "loss": 0.5302,
+      "step": 366
+    },
+    {
+      "epoch": 1.1314065510597302,
+      "grad_norm": 0.6757516012055022,
+      "learning_rate": 7.61607232238443e-05,
+      "loss": 0.5351,
+      "step": 367
+    },
+    {
+      "epoch": 1.1344894026974952,
+      "grad_norm": 1.0059480136856733,
+      "learning_rate": 7.612379396619748e-05,
+      "loss": 0.5435,
+      "step": 368
+    },
+    {
+      "epoch": 1.13757225433526,
+      "grad_norm": 1.1953008272895773,
+      "learning_rate": 7.608669699124153e-05,
+      "loss": 0.5388,
+      "step": 369
+    },
+    {
+      "epoch": 1.140655105973025,
+      "grad_norm": 1.0180012961868927,
+      "learning_rate": 7.604943247121203e-05,
+      "loss": 0.5361,
+      "step": 370
+    },
+    {
+      "epoch": 1.14373795761079,
+      "grad_norm": 1.018565841052241,
+      "learning_rate": 7.601200057912257e-05,
+      "loss": 0.5394,
+      "step": 371
+    },
+    {
+      "epoch": 1.146820809248555,
+      "grad_norm": 0.938166709551276,
+      "learning_rate": 7.597440148876373e-05,
+      "loss": 0.5407,
+      "step": 372
+    },
+    {
+      "epoch": 1.1499036608863198,
+      "grad_norm": 0.8756784667189219,
+      "learning_rate": 7.593663537470241e-05,
+      "loss": 0.5372,
+      "step": 373
+    },
+    {
+      "epoch": 1.1529865125240848,
+      "grad_norm": 0.8129522524062,
+      "learning_rate": 7.589870241228095e-05,
+      "loss": 0.5368,
+      "step": 374
+    },
+    {
+      "epoch": 1.1560693641618498,
+      "grad_norm": 0.8516986016789037,
+      "learning_rate": 7.586060277761638e-05,
+      "loss": 0.534,
+      "step": 375
+    },
+    {
+      "epoch": 1.1591522157996146,
+      "grad_norm": 0.768620760847617,
+      "learning_rate": 7.58223366475995e-05,
+      "loss": 0.538,
+      "step": 376
+    },
+    {
+      "epoch": 1.1622350674373796,
+      "grad_norm": 0.6512017766341572,
+      "learning_rate": 7.57839041998942e-05,
+      "loss": 0.5351,
+      "step": 377
+    },
+    {
+      "epoch": 1.1653179190751446,
+      "grad_norm": 0.48539567154100205,
+      "learning_rate": 7.57453056129365e-05,
+      "loss": 0.5382,
+      "step": 378
+    },
+    {
+      "epoch": 1.1684007707129094,
+      "grad_norm": 0.6845213618402091,
+      "learning_rate": 7.570654106593379e-05,
+      "loss": 0.5329,
+      "step": 379
+    },
+    {
+      "epoch": 1.1714836223506744,
+      "grad_norm": 0.750525953378347,
+      "learning_rate": 7.566761073886402e-05,
+      "loss": 0.5291,
+      "step": 380
+    },
+    {
+      "epoch": 1.1745664739884394,
+      "grad_norm": 0.6726675615828184,
+      "learning_rate": 7.562851481247478e-05,
+      "loss": 0.5402,
+      "step": 381
+    },
+    {
+      "epoch": 1.1776493256262042,
+      "grad_norm": 0.6736976519780983,
+      "learning_rate": 7.558925346828256e-05,
+      "loss": 0.5313,
+      "step": 382
+    },
+    {
+      "epoch": 1.1807321772639692,
+      "grad_norm": 0.7064067369853219,
+      "learning_rate": 7.554982688857183e-05,
+      "loss": 0.5336,
+      "step": 383
+    },
+    {
+      "epoch": 1.1838150289017342,
+      "grad_norm": 0.640407455585253,
+      "learning_rate": 7.551023525639425e-05,
+      "loss": 0.5332,
+      "step": 384
+    },
+    {
+      "epoch": 1.186897880539499,
+      "grad_norm": 0.5858330999812315,
+      "learning_rate": 7.547047875556775e-05,
+      "loss": 0.5328,
+      "step": 385
+    },
+    {
+      "epoch": 1.189980732177264,
+      "grad_norm": 0.6219088549116051,
+      "learning_rate": 7.543055757067576e-05,
+      "loss": 0.5311,
+      "step": 386
+    },
+    {
+      "epoch": 1.193063583815029,
+      "grad_norm": 0.637344161103186,
+      "learning_rate": 7.539047188706631e-05,
+      "loss": 0.5326,
+      "step": 387
+    },
+    {
+      "epoch": 1.1961464354527938,
+      "grad_norm": 0.8550099379650122,
+      "learning_rate": 7.535022189085115e-05,
+      "loss": 0.5338,
+      "step": 388
+    },
+    {
+      "epoch": 1.1992292870905588,
+      "grad_norm": 1.1623707424458203,
+      "learning_rate": 7.530980776890495e-05,
+      "loss": 0.532,
+      "step": 389
+    },
+    {
+      "epoch": 1.2023121387283238,
+      "grad_norm": 1.0536433449057823,
+      "learning_rate": 7.526922970886432e-05,
+      "loss": 0.5385,
+      "step": 390
+    },
+    {
+      "epoch": 1.2053949903660885,
+      "grad_norm": 0.9018309013378271,
+      "learning_rate": 7.522848789912708e-05,
+      "loss": 0.5348,
+      "step": 391
+    },
+    {
+      "epoch": 1.2084778420038536,
+      "grad_norm": 0.9151308724871474,
+      "learning_rate": 7.518758252885128e-05,
+      "loss": 0.5341,
+      "step": 392
+    },
+    {
+      "epoch": 1.2115606936416186,
+      "grad_norm": 1.0795148162541461,
+      "learning_rate": 7.514651378795438e-05,
+      "loss": 0.5316,
+      "step": 393
+    },
+    {
+      "epoch": 1.2146435452793833,
+      "grad_norm": 1.0058467558765545,
+      "learning_rate": 7.510528186711229e-05,
+      "loss": 0.5334,
+      "step": 394
+    },
+    {
+      "epoch": 1.2177263969171483,
+      "grad_norm": 0.9191357696142056,
+      "learning_rate": 7.50638869577586e-05,
+      "loss": 0.5341,
+      "step": 395
+    },
+    {
+      "epoch": 1.2208092485549134,
+      "grad_norm": 0.8458265076401587,
+      "learning_rate": 7.502232925208365e-05,
+      "loss": 0.5264,
+      "step": 396
+    },
+    {
+      "epoch": 1.2238921001926781,
+      "grad_norm": 0.949215922322586,
+      "learning_rate": 7.498060894303351e-05,
+      "loss": 0.5346,
+      "step": 397
+    },
+    {
+      "epoch": 1.2269749518304431,
+      "grad_norm": 1.1766025971055396,
+      "learning_rate": 7.49387262243093e-05,
+      "loss": 0.5335,
+      "step": 398
+    },
+    {
+      "epoch": 1.2300578034682081,
+      "grad_norm": 0.7186332324947287,
+      "learning_rate": 7.489668129036615e-05,
+      "loss": 0.5317,
+      "step": 399
+    },
+    {
+      "epoch": 1.2331406551059731,
+      "grad_norm": 0.43493000210784305,
+      "learning_rate": 7.485447433641232e-05,
+      "loss": 0.5327,
+      "step": 400
+    },
+    {
+      "epoch": 1.236223506743738,
+      "grad_norm": 0.6125292612546878,
+      "learning_rate": 7.48121055584083e-05,
+      "loss": 0.5332,
+      "step": 401
+    },
+    {
+      "epoch": 1.239306358381503,
+      "grad_norm": 0.797620883058671,
+      "learning_rate": 7.476957515306591e-05,
+      "loss": 0.5305,
+      "step": 402
+    },
+    {
+      "epoch": 1.242389210019268,
+      "grad_norm": 0.8613821459845584,
+      "learning_rate": 7.472688331784742e-05,
+      "loss": 0.5318,
+      "step": 403
+    },
+    {
+      "epoch": 1.2454720616570327,
+      "grad_norm": 0.8170389987033159,
+      "learning_rate": 7.468403025096453e-05,
+      "loss": 0.5273,
+      "step": 404
+    },
+    {
+      "epoch": 1.2485549132947977,
+      "grad_norm": 0.7812936637027204,
+      "learning_rate": 7.464101615137756e-05,
+      "loss": 0.5309,
+      "step": 405
+    },
+    {
+      "epoch": 1.2516377649325627,
+      "grad_norm": 0.9003967234997579,
+      "learning_rate": 7.459784121879444e-05,
+      "loss": 0.5365,
+      "step": 406
+    },
+    {
+      "epoch": 1.2547206165703275,
+      "grad_norm": 0.965571382834061,
+      "learning_rate": 7.455450565366989e-05,
+      "loss": 0.528,
+      "step": 407
+    },
+    {
+      "epoch": 1.2578034682080925,
+      "grad_norm": 1.0338715985620017,
+      "learning_rate": 7.451100965720435e-05,
+      "loss": 0.5284,
+      "step": 408
+    },
+    {
+      "epoch": 1.2608863198458575,
+      "grad_norm": 1.148860989789251,
+      "learning_rate": 7.446735343134317e-05,
+      "loss": 0.5337,
+      "step": 409
+    },
+    {
+      "epoch": 1.2639691714836223,
+      "grad_norm": 0.828323685982287,
+      "learning_rate": 7.442353717877557e-05,
+      "loss": 0.5304,
+      "step": 410
+    },
+    {
+      "epoch": 1.2670520231213873,
+      "grad_norm": 0.5824577770488141,
+      "learning_rate": 7.437956110293381e-05,
+      "loss": 0.529,
+      "step": 411
+    },
+    {
+      "epoch": 1.2701348747591523,
+      "grad_norm": 0.5756974153483788,
+      "learning_rate": 7.433542540799214e-05,
+      "loss": 0.5287,
+      "step": 412
+    },
+    {
+      "epoch": 1.273217726396917,
+      "grad_norm": 0.6839597944908083,
+      "learning_rate": 7.429113029886595e-05,
+      "loss": 0.5304,
+      "step": 413
+    },
+    {
+      "epoch": 1.276300578034682,
+      "grad_norm": 0.8815936835044538,
+      "learning_rate": 7.424667598121067e-05,
+      "loss": 0.5268,
+      "step": 414
+    },
+    {
+      "epoch": 1.2793834296724471,
+      "grad_norm": 1.1800895082303688,
+      "learning_rate": 7.420206266142099e-05,
+      "loss": 0.5351,
+      "step": 415
+    },
+    {
+      "epoch": 1.282466281310212,
+      "grad_norm": 0.7719006481113085,
+      "learning_rate": 7.415729054662983e-05,
+      "loss": 0.5349,
+      "step": 416
+    },
+    {
+      "epoch": 1.285549132947977,
+      "grad_norm": 0.5081761530511858,
+      "learning_rate": 7.411235984470732e-05,
+      "loss": 0.5262,
+      "step": 417
+    },
+    {
+      "epoch": 1.288631984585742,
+      "grad_norm": 0.6291321319152572,
+      "learning_rate": 7.406727076425992e-05,
+      "loss": 0.531,
+      "step": 418
+    },
+    {
+      "epoch": 1.2917148362235067,
+      "grad_norm": 0.8854219802912666,
+      "learning_rate": 7.40220235146294e-05,
+      "loss": 0.5319,
+      "step": 419
+    },
+    {
+      "epoch": 1.2947976878612717,
+      "grad_norm": 1.0086518516460292,
+      "learning_rate": 7.39766183058919e-05,
+      "loss": 0.5336,
+      "step": 420
+    },
+    {
+      "epoch": 1.2978805394990367,
+      "grad_norm": 1.1830895036689468,
+      "learning_rate": 7.393105534885692e-05,
+      "loss": 0.5326,
+      "step": 421
+    },
+    {
+      "epoch": 1.3009633911368015,
+      "grad_norm": 0.818661670387812,
+      "learning_rate": 7.388533485506639e-05,
+      "loss": 0.5294,
+      "step": 422
+    },
+    {
+      "epoch": 1.3040462427745665,
+      "grad_norm": 0.45345778403222053,
+      "learning_rate": 7.383945703679365e-05,
+      "loss": 0.5282,
+      "step": 423
+    },
+    {
+      "epoch": 1.3071290944123315,
+      "grad_norm": 0.6524682531195315,
+      "learning_rate": 7.379342210704245e-05,
+      "loss": 0.5298,
+      "step": 424
+    },
+    {
+      "epoch": 1.3102119460500963,
+      "grad_norm": 1.1053419915010056,
+      "learning_rate": 7.3747230279546e-05,
+      "loss": 0.5363,
+      "step": 425
+    },
+    {
+      "epoch": 1.3132947976878613,
+      "grad_norm": 1.0113484237143457,
+      "learning_rate": 7.370088176876598e-05,
+      "loss": 0.5254,
+      "step": 426
+    },
+    {
+      "epoch": 1.3163776493256263,
+      "grad_norm": 0.8688583428029235,
+      "learning_rate": 7.365437678989152e-05,
+      "loss": 0.5301,
+      "step": 427
+    },
+    {
+      "epoch": 1.319460500963391,
+      "grad_norm": 0.8740799511780297,
+      "learning_rate": 7.360771555883818e-05,
+      "loss": 0.535,
+      "step": 428
+    },
+    {
+      "epoch": 1.322543352601156,
+      "grad_norm": 1.0280790795350367,
+      "learning_rate": 7.356089829224699e-05,
+      "loss": 0.5306,
+      "step": 429
+    },
+    {
+      "epoch": 1.325626204238921,
+      "grad_norm": 1.0869540127718802,
+      "learning_rate": 7.351392520748346e-05,
+      "loss": 0.5306,
+      "step": 430
+    },
+    {
+      "epoch": 1.3287090558766859,
+      "grad_norm": 0.9495744427345415,
+      "learning_rate": 7.34667965226365e-05,
+      "loss": 0.5324,
+      "step": 431
+    },
+    {
+      "epoch": 1.3317919075144509,
+      "grad_norm": 0.8316374989805042,
+      "learning_rate": 7.341951245651747e-05,
+      "loss": 0.529,
+      "step": 432
+    },
+    {
+      "epoch": 1.3348747591522159,
+      "grad_norm": 0.5794576671173408,
+      "learning_rate": 7.337207322865911e-05,
+      "loss": 0.5272,
+      "step": 433
+    },
+    {
+      "epoch": 1.3379576107899807,
+      "grad_norm": 0.4441102435668441,
+      "learning_rate": 7.33244790593146e-05,
+      "loss": 0.5269,
+      "step": 434
+    },
+    {
+      "epoch": 1.3410404624277457,
+      "grad_norm": 0.5580717845135446,
+      "learning_rate": 7.327673016945647e-05,
+      "loss": 0.5295,
+      "step": 435
+    },
+    {
+      "epoch": 1.3441233140655107,
+      "grad_norm": 0.6647670035168102,
+      "learning_rate": 7.322882678077557e-05,
+      "loss": 0.5306,
+      "step": 436
+    },
+    {
+      "epoch": 1.3472061657032754,
+      "grad_norm": 0.6892997442618053,
+      "learning_rate": 7.318076911568011e-05,
+      "loss": 0.5296,
+      "step": 437
+    },
+    {
+      "epoch": 1.3502890173410405,
+      "grad_norm": 0.653779776759592,
+      "learning_rate": 7.313255739729453e-05,
+      "loss": 0.5285,
+      "step": 438
+    },
+    {
+      "epoch": 1.3533718689788055,
+      "grad_norm": 0.5916605739047195,
+      "learning_rate": 7.308419184945855e-05,
+      "loss": 0.5311,
+      "step": 439
+    },
+    {
+      "epoch": 1.3564547206165702,
+      "grad_norm": 0.48221532470648926,
+      "learning_rate": 7.303567269672612e-05,
+      "loss": 0.5291,
+      "step": 440
+    },
+    {
+      "epoch": 1.3595375722543352,
+      "grad_norm": 0.40169932172272543,
+      "learning_rate": 7.298700016436427e-05,
+      "loss": 0.5281,
+      "step": 441
+    },
+    {
+      "epoch": 1.3626204238921003,
+      "grad_norm": 0.5507733044328121,
+      "learning_rate": 7.293817447835226e-05,
+      "loss": 0.5299,
+      "step": 442
+    },
+    {
+      "epoch": 1.365703275529865,
+      "grad_norm": 0.6707410460073495,
+      "learning_rate": 7.288919586538031e-05,
+      "loss": 0.5294,
+      "step": 443
+    },
+    {
+      "epoch": 1.36878612716763,
+      "grad_norm": 0.8326244528817593,
+      "learning_rate": 7.284006455284874e-05,
+      "loss": 0.5322,
+      "step": 444
+    },
+    {
+      "epoch": 1.371868978805395,
+      "grad_norm": 1.0333382607633659,
+      "learning_rate": 7.279078076886681e-05,
+      "loss": 0.5323,
+      "step": 445
+    },
+    {
+      "epoch": 1.3749518304431598,
+      "grad_norm": 0.8909496896249056,
+      "learning_rate": 7.274134474225163e-05,
+      "loss": 0.5349,
+      "step": 446
+    },
+    {
+      "epoch": 1.3780346820809248,
+      "grad_norm": 0.6832570359328524,
+      "learning_rate": 7.269175670252723e-05,
+      "loss": 0.5295,
+      "step": 447
+    },
+    {
+      "epoch": 1.3811175337186898,
+      "grad_norm": 0.7241740754505229,
+      "learning_rate": 7.264201687992336e-05,
+      "loss": 0.5321,
+      "step": 448
+    },
+    {
+      "epoch": 1.3842003853564546,
+      "grad_norm": 0.8618486669507248,
+      "learning_rate": 7.259212550537448e-05,
+      "loss": 0.5282,
+      "step": 449
+    },
+    {
+      "epoch": 1.3872832369942196,
+      "grad_norm": 0.9246035069730334,
+      "learning_rate": 7.254208281051871e-05,
+      "loss": 0.5239,
+      "step": 450
+    },
+    {
+      "epoch": 1.3903660886319846,
+      "grad_norm": 1.0298643019766507,
+      "learning_rate": 7.24918890276967e-05,
+      "loss": 0.5293,
+      "step": 451
+    },
+    {
+      "epoch": 1.3934489402697494,
+      "grad_norm": 1.037223032239434,
+      "learning_rate": 7.244154438995059e-05,
+      "loss": 0.5256,
+      "step": 452
+    },
+    {
+      "epoch": 1.3965317919075144,
+      "grad_norm": 0.9372040375501816,
+      "learning_rate": 7.23910491310229e-05,
+      "loss": 0.5249,
+      "step": 453
+    },
+    {
+      "epoch": 1.3996146435452794,
+      "grad_norm": 1.0228412458440592,
+      "learning_rate": 7.234040348535552e-05,
+      "loss": 0.5308,
+      "step": 454
+    },
+    {
+      "epoch": 1.4026974951830442,
+      "grad_norm": 1.1408235726993474,
+      "learning_rate": 7.228960768808847e-05,
+      "loss": 0.5228,
+      "step": 455
+    },
+    {
+      "epoch": 1.4057803468208092,
+      "grad_norm": 0.5840814968829888,
+      "learning_rate": 7.223866197505898e-05,
+      "loss": 0.5278,
+      "step": 456
+    },
+    {
+      "epoch": 1.4088631984585742,
+      "grad_norm": 0.5431403323578895,
+      "learning_rate": 7.218756658280028e-05,
+      "loss": 0.5251,
+      "step": 457
+    },
+    {
+      "epoch": 1.411946050096339,
+      "grad_norm": 0.7279985775643657,
+      "learning_rate": 7.213632174854052e-05,
+      "loss": 0.5258,
+      "step": 458
+    },
+    {
+      "epoch": 1.415028901734104,
+      "grad_norm": 0.7456319636554177,
+      "learning_rate": 7.208492771020176e-05,
+      "loss": 0.5282,
+      "step": 459
+    },
+    {
+      "epoch": 1.418111753371869,
+      "grad_norm": 0.7681779695441887,
+      "learning_rate": 7.20333847063987e-05,
+      "loss": 0.5299,
+      "step": 460
+    },
+    {
+      "epoch": 1.4211946050096338,
+      "grad_norm": 0.7056739945801672,
+      "learning_rate": 7.198169297643774e-05,
+      "loss": 0.53,
+      "step": 461
+    },
+    {
+      "epoch": 1.4242774566473988,
+      "grad_norm": 0.6199592602135798,
+      "learning_rate": 7.192985276031571e-05,
+      "loss": 0.5295,
+      "step": 462
+    },
+    {
+      "epoch": 1.4273603082851638,
+      "grad_norm": 0.5330894743305711,
+      "learning_rate": 7.187786429871896e-05,
+      "loss": 0.5262,
+      "step": 463
+    },
+    {
+      "epoch": 1.4304431599229286,
+      "grad_norm": 0.4538305818015674,
+      "learning_rate": 7.1825727833022e-05,
+      "loss": 0.5288,
+      "step": 464
+    },
+    {
+      "epoch": 1.4335260115606936,
+      "grad_norm": 0.34173497587451784,
+      "learning_rate": 7.177344360528656e-05,
+      "loss": 0.5253,
+      "step": 465
+    },
+    {
+      "epoch": 1.4366088631984586,
+      "grad_norm": 0.35045179429364176,
+      "learning_rate": 7.17210118582604e-05,
+      "loss": 0.524,
+      "step": 466
+    },
+    {
+      "epoch": 1.4396917148362234,
+      "grad_norm": 0.4322920758280479,
+      "learning_rate": 7.166843283537619e-05,
+      "loss": 0.5232,
+      "step": 467
+    },
+    {
+      "epoch": 1.4427745664739884,
+      "grad_norm": 0.445424985824893,
+      "learning_rate": 7.161570678075038e-05,
+      "loss": 0.5281,
+      "step": 468
+    },
+    {
+      "epoch": 1.4458574181117534,
+      "grad_norm": 0.4363959598652878,
+      "learning_rate": 7.156283393918205e-05,
+      "loss": 0.5267,
+      "step": 469
+    },
+    {
+      "epoch": 1.4489402697495182,
+      "grad_norm": 0.42031382065851414,
+      "learning_rate": 7.150981455615183e-05,
+      "loss": 0.5247,
+      "step": 470
+    },
+    {
+      "epoch": 1.4520231213872832,
+      "grad_norm": 0.44263091904119944,
+      "learning_rate": 7.145664887782066e-05,
+      "loss": 0.5239,
+      "step": 471
+    },
+    {
+      "epoch": 1.4551059730250482,
+      "grad_norm": 0.4988065487930374,
+      "learning_rate": 7.140333715102876e-05,
+      "loss": 0.5237,
+      "step": 472
+    },
+    {
+      "epoch": 1.458188824662813,
+      "grad_norm": 0.6792933047699445,
+      "learning_rate": 7.134987962329441e-05,
+      "loss": 0.5308,
+      "step": 473
+    },
+    {
+      "epoch": 1.461271676300578,
+      "grad_norm": 0.9880727368094488,
+      "learning_rate": 7.12962765428128e-05,
+      "loss": 0.53,
+      "step": 474
+    },
+    {
+      "epoch": 1.464354527938343,
+      "grad_norm": 1.2528860369347115,
+      "learning_rate": 7.124252815845494e-05,
+      "loss": 0.5308,
+      "step": 475
+    },
+    {
+      "epoch": 1.467437379576108,
+      "grad_norm": 0.6828724893215994,
+      "learning_rate": 7.118863471976645e-05,
+      "loss": 0.5286,
+      "step": 476
+    },
+    {
+      "epoch": 1.4705202312138728,
+      "grad_norm": 0.6819515540386075,
+      "learning_rate": 7.113459647696641e-05,
+      "loss": 0.5243,
+      "step": 477
+    },
+    {
+      "epoch": 1.4736030828516378,
+      "grad_norm": 0.8858104592391481,
+      "learning_rate": 7.108041368094619e-05,
+      "loss": 0.5235,
+      "step": 478
+    },
+    {
+      "epoch": 1.4766859344894028,
+      "grad_norm": 1.0702213255688144,
+      "learning_rate": 7.102608658326831e-05,
+      "loss": 0.527,
+      "step": 479
+    },
+    {
+      "epoch": 1.4797687861271676,
+      "grad_norm": 1.035198328947949,
+      "learning_rate": 7.09716154361653e-05,
+      "loss": 0.5265,
+      "step": 480
+    },
+    {
+      "epoch": 1.4828516377649326,
+      "grad_norm": 1.067928650234488,
+      "learning_rate": 7.091700049253842e-05,
+      "loss": 0.525,
+      "step": 481
+    },
+    {
+      "epoch": 1.4859344894026976,
+      "grad_norm": 0.9939995816472853,
+      "learning_rate": 7.086224200595659e-05,
+      "loss": 0.5317,
+      "step": 482
+    },
+    {
+      "epoch": 1.4890173410404623,
+      "grad_norm": 0.9912810284665426,
+      "learning_rate": 7.08073402306552e-05,
+      "loss": 0.5338,
+      "step": 483
+    },
+    {
+      "epoch": 1.4921001926782274,
+      "grad_norm": 0.9485054245374028,
+      "learning_rate": 7.075229542153485e-05,
+      "loss": 0.5295,
+      "step": 484
+    },
+    {
+      "epoch": 1.4951830443159924,
+      "grad_norm": 0.9218883984974872,
+      "learning_rate": 7.069710783416029e-05,
+      "loss": 0.5262,
+      "step": 485
+    },
+    {
+      "epoch": 1.4982658959537571,
+      "grad_norm": 0.9139559810929375,
+      "learning_rate": 7.064177772475912e-05,
+      "loss": 0.5288,
+      "step": 486
+    },
+    {
+      "epoch": 1.5013487475915221,
+      "grad_norm": 0.7739799607410872,
+      "learning_rate": 7.058630535022069e-05,
+      "loss": 0.5234,
+      "step": 487
+    },
+    {
+      "epoch": 1.504431599229287,
+      "grad_norm": 0.5925891464635009,
+      "learning_rate": 7.053069096809481e-05,
+      "loss": 0.5237,
+      "step": 488
+    },
+    {
+      "epoch": 1.5075144508670522,
+      "grad_norm": 0.5474148096828393,
+      "learning_rate": 7.047493483659064e-05,
+      "loss": 0.5237,
+      "step": 489
+    },
+    {
+      "epoch": 1.510597302504817,
+      "grad_norm": 0.6665713472337279,
+      "learning_rate": 7.04190372145755e-05,
+      "loss": 0.5231,
+      "step": 490
+    },
+    {
+      "epoch": 1.5136801541425817,
+      "grad_norm": 0.7310284629993853,
+      "learning_rate": 7.036299836157355e-05,
+      "loss": 0.5291,
+      "step": 491
+    },
+    {
+      "epoch": 1.516763005780347,
+      "grad_norm": 0.5557357453870334,
+      "learning_rate": 7.030681853776472e-05,
+      "loss": 0.523,
+      "step": 492
+    },
+    {
+      "epoch": 1.5198458574181117,
+      "grad_norm": 0.5544710517728318,
+      "learning_rate": 7.025049800398344e-05,
+      "loss": 0.528,
+      "step": 493
+    },
+    {
+      "epoch": 1.5229287090558767,
+      "grad_norm": 0.6960721731676911,
+      "learning_rate": 7.019403702171741e-05,
+      "loss": 0.5272,
+      "step": 494
+    },
+    {
+      "epoch": 1.5260115606936417,
+      "grad_norm": 0.5577721288828663,
+      "learning_rate": 7.013743585310642e-05,
+      "loss": 0.5228,
+      "step": 495
+    },
+    {
+      "epoch": 1.5290944123314065,
+      "grad_norm": 0.5666596305386035,
+      "learning_rate": 7.008069476094118e-05,
+      "loss": 0.5223,
+      "step": 496
+    },
+    {
+      "epoch": 1.5321772639691715,
+      "grad_norm": 0.6513194148959046,
+      "learning_rate": 7.002381400866194e-05,
+      "loss": 0.5228,
+      "step": 497
+    },
+    {
+      "epoch": 1.5352601156069365,
+      "grad_norm": 0.8870935548702494,
+      "learning_rate": 6.996679386035744e-05,
+      "loss": 0.5278,
+      "step": 498
+    },
+    {
+      "epoch": 1.5383429672447013,
+      "grad_norm": 0.943714541750328,
+      "learning_rate": 6.990963458076363e-05,
+      "loss": 0.5224,
+      "step": 499
+    },
+    {
+      "epoch": 1.5414258188824663,
+      "grad_norm": 0.8168206392811749,
+      "learning_rate": 6.985233643526235e-05,
+      "loss": 0.5217,
+      "step": 500
+    },
+    {
+      "epoch": 1.5445086705202313,
+      "grad_norm": 0.7095406841155505,
+      "learning_rate": 6.979489968988024e-05,
+      "loss": 0.5214,
+      "step": 501
+    },
+    {
+      "epoch": 1.5475915221579961,
+      "grad_norm": 0.7600951388889429,
+      "learning_rate": 6.973732461128743e-05,
+      "loss": 0.527,
+      "step": 502
+    },
+    {
+      "epoch": 1.5506743737957611,
+      "grad_norm": 0.8589830764412946,
+      "learning_rate": 6.967961146679629e-05,
+      "loss": 0.5246,
+      "step": 503
+    },
+    {
+      "epoch": 1.5537572254335261,
+      "grad_norm": 0.9097390198825365,
+      "learning_rate": 6.96217605243602e-05,
+      "loss": 0.5275,
+      "step": 504
+    },
+    {
+      "epoch": 1.556840077071291,
+      "grad_norm": 0.912572270049121,
+      "learning_rate": 6.956377205257234e-05,
+      "loss": 0.5234,
+      "step": 505
+    },
+    {
+      "epoch": 1.559922928709056,
+      "grad_norm": 0.8290359084267758,
+      "learning_rate": 6.950564632066442e-05,
+      "loss": 0.5251,
+      "step": 506
+    },
+    {
+      "epoch": 1.563005780346821,
+      "grad_norm": 0.8213443660474333,
+      "learning_rate": 6.944738359850542e-05,
+      "loss": 0.5266,
+      "step": 507
+    },
+    {
+      "epoch": 1.5660886319845857,
+      "grad_norm": 0.8208861205731054,
+      "learning_rate": 6.938898415660034e-05,
+      "loss": 0.5242,
+      "step": 508
+    },
+    {
+      "epoch": 1.5691714836223507,
+      "grad_norm": 0.912752747462451,
+      "learning_rate": 6.933044826608892e-05,
+      "loss": 0.5216,
+      "step": 509
+    },
+    {
+      "epoch": 1.5722543352601157,
+      "grad_norm": 1.1353586022806916,
+      "learning_rate": 6.92717761987445e-05,
+      "loss": 0.5231,
+      "step": 510
+    },
+    {
+      "epoch": 1.5753371868978805,
+      "grad_norm": 7.174349988008314,
+      "learning_rate": 6.921296822697256e-05,
+      "loss": 0.5283,
+      "step": 511
+    },
+    {
+      "epoch": 1.5784200385356455,
+      "grad_norm": 1.8911600440181702,
+      "learning_rate": 6.915402462380963e-05,
+      "loss": 0.528,
+      "step": 512
+    },
+    {
+      "epoch": 1.5815028901734105,
+      "grad_norm": 0.47329137898505325,
+      "learning_rate": 6.909494566292195e-05,
+      "loss": 0.5217,
+      "step": 513
+    },
+    {
+      "epoch": 1.5845857418111753,
+      "grad_norm": 1.1313832415881564,
+      "learning_rate": 6.903573161860419e-05,
+      "loss": 0.5269,
+      "step": 514
+    },
+    {
+      "epoch": 1.5876685934489403,
+      "grad_norm": 1.3036915769693034,
+      "learning_rate": 6.897638276577819e-05,
+      "loss": 0.5228,
+      "step": 515
+    },
+    {
+      "epoch": 1.5907514450867053,
+      "grad_norm": 0.4093565863948932,
+      "learning_rate": 6.891689937999172e-05,
+      "loss": 0.5211,
+      "step": 516
+    },
+    {
+      "epoch": 1.59383429672447,
+      "grad_norm": 1.178435990591182,
+      "learning_rate": 6.88572817374171e-05,
+      "loss": 0.5249,
+      "step": 517
+    },
+    {
+      "epoch": 1.596917148362235,
+      "grad_norm": 0.8880736123149384,
+      "learning_rate": 6.879753011485006e-05,
+      "loss": 0.5249,
+      "step": 518
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6235785732030509,
+      "learning_rate": 6.873764478970831e-05,
+      "loss": 0.5236,
+      "step": 519
+    },
+    {
+      "epoch": 1.6030828516377649,
+      "grad_norm": 0.5264930021043772,
+      "learning_rate": 6.867762604003037e-05,
+      "loss": 0.5218,
+      "step": 520
+    },
+    {
+      "epoch": 1.6061657032755299,
+      "grad_norm": 0.4523005553245844,
+      "learning_rate": 6.861747414447421e-05,
+      "loss": 0.524,
+      "step": 521
+    },
+    {
+      "epoch": 1.6092485549132949,
+      "grad_norm": 0.4785784811617481,
+      "learning_rate": 6.855718938231597e-05,
+      "loss": 0.5212,
+      "step": 522
+    },
+    {
+      "epoch": 1.6123314065510597,
+      "grad_norm": 0.5268203599243676,
+      "learning_rate": 6.849677203344869e-05,
+      "loss": 0.5222,
+      "step": 523
+    },
+    {
+      "epoch": 1.6154142581888247,
+      "grad_norm": 0.4794091535300212,
+      "learning_rate": 6.843622237838098e-05,
+      "loss": 0.5231,
+      "step": 524
+    },
+    {
+      "epoch": 1.6184971098265897,
+      "grad_norm": 0.4820717969180606,
+      "learning_rate": 6.837554069823572e-05,
+      "loss": 0.5207,
+      "step": 525
+    },
+    {
+      "epoch": 1.6215799614643545,
+      "grad_norm": 0.5570363607644362,
+      "learning_rate": 6.831472727474878e-05,
+      "loss": 0.5231,
+      "step": 526
+    },
+    {
+      "epoch": 1.6246628131021195,
+      "grad_norm": 0.5405930175270516,
+      "learning_rate": 6.825378239026766e-05,
+      "loss": 0.5245,
+      "step": 527
+    },
+    {
+      "epoch": 1.6277456647398845,
+      "grad_norm": 0.5417762417585796,
+      "learning_rate": 6.819270632775025e-05,
+      "loss": 0.5218,
+      "step": 528
+    },
+    {
+      "epoch": 1.6308285163776493,
+      "grad_norm": 0.6225042500849538,
+      "learning_rate": 6.813149937076347e-05,
+      "loss": 0.522,
+      "step": 529
+    },
+    {
+      "epoch": 1.6339113680154143,
+      "grad_norm": 0.6916050459483382,
+      "learning_rate": 6.807016180348195e-05,
+      "loss": 0.5265,
+      "step": 530
+    },
+    {
+      "epoch": 1.6369942196531793,
+      "grad_norm": 0.5967493529148962,
+      "learning_rate": 6.800869391068674e-05,
+      "loss": 0.5165,
+      "step": 531
+    },
+    {
+      "epoch": 1.640077071290944,
+      "grad_norm": 0.5378537474139249,
+      "learning_rate": 6.794709597776398e-05,
+      "loss": 0.5236,
+      "step": 532
+    },
+    {
+      "epoch": 1.643159922928709,
+      "grad_norm": 0.6931598671521784,
+      "learning_rate": 6.788536829070351e-05,
+      "loss": 0.5233,
+      "step": 533
+    },
+    {
+      "epoch": 1.646242774566474,
+      "grad_norm": 0.9612453498911665,
+      "learning_rate": 6.782351113609767e-05,
+      "loss": 0.5181,
+      "step": 534
+    },
+    {
+      "epoch": 1.6493256262042388,
+      "grad_norm": 1.2737503401084873,
+      "learning_rate": 6.776152480113989e-05,
+      "loss": 0.5228,
+      "step": 535
+    },
+    {
+      "epoch": 1.6524084778420038,
+      "grad_norm": 0.6398735834455941,
+      "learning_rate": 6.76994095736233e-05,
+      "loss": 0.5212,
+      "step": 536
+    },
+    {
+      "epoch": 1.6554913294797688,
+      "grad_norm": 0.46779792854692176,
+      "learning_rate": 6.763716574193952e-05,
+      "loss": 0.5257,
+      "step": 537
+    },
+    {
+      "epoch": 1.6585741811175336,
+      "grad_norm": 0.6530745395320384,
+      "learning_rate": 6.757479359507724e-05,
+      "loss": 0.5175,
+      "step": 538
+    },
+    {
+      "epoch": 1.6616570327552986,
+      "grad_norm": 0.9127472185837223,
+      "learning_rate": 6.75122934226209e-05,
+      "loss": 0.5251,
+      "step": 539
+    },
+    {
+      "epoch": 1.6647398843930636,
+      "grad_norm": 1.0848306395457255,
+      "learning_rate": 6.744966551474936e-05,
+      "loss": 0.5266,
+      "step": 540
+    },
+    {
+      "epoch": 1.6678227360308284,
+      "grad_norm": 0.7890691640189261,
+      "learning_rate": 6.738691016223449e-05,
+      "loss": 0.5219,
+      "step": 541
+    },
+    {
+      "epoch": 1.6709055876685934,
+      "grad_norm": 0.57469503694783,
+      "learning_rate": 6.73240276564399e-05,
+      "loss": 0.5255,
+      "step": 542
+    },
+    {
+      "epoch": 1.6739884393063584,
+      "grad_norm": 0.5291533455080574,
+      "learning_rate": 6.726101828931958e-05,
+      "loss": 0.5216,
+      "step": 543
+    },
+    {
+      "epoch": 1.6770712909441232,
+      "grad_norm": 0.5884554890143913,
+      "learning_rate": 6.719788235341645e-05,
+      "loss": 0.5217,
+      "step": 544
+    },
+    {
+      "epoch": 1.6801541425818882,
+      "grad_norm": 0.6370708641487587,
+      "learning_rate": 6.713462014186113e-05,
+      "loss": 0.5223,
+      "step": 545
+    },
+    {
+      "epoch": 1.6832369942196532,
+      "grad_norm": 0.6423011706165014,
+      "learning_rate": 6.707123194837049e-05,
+      "loss": 0.5206,
+      "step": 546
+    },
+    {
+      "epoch": 1.686319845857418,
+      "grad_norm": 0.5818838027124593,
+      "learning_rate": 6.70077180672463e-05,
+      "loss": 0.5228,
+      "step": 547
+    },
+    {
+      "epoch": 1.689402697495183,
+      "grad_norm": 0.5107044521315691,
+      "learning_rate": 6.694407879337394e-05,
+      "loss": 0.5183,
+      "step": 548
+    },
+    {
+      "epoch": 1.692485549132948,
+      "grad_norm": 0.5790100685013165,
+      "learning_rate": 6.688031442222091e-05,
+      "loss": 0.518,
+      "step": 549
+    },
+    {
+      "epoch": 1.6955684007707128,
+      "grad_norm": 0.6647380024495569,
+      "learning_rate": 6.681642524983551e-05,
+      "loss": 0.5191,
+      "step": 550
+    },
+    {
+      "epoch": 1.6986512524084778,
+      "grad_norm": 0.6036358466724143,
+      "learning_rate": 6.675241157284553e-05,
+      "loss": 0.5189,
+      "step": 551
+    },
+    {
+      "epoch": 1.7017341040462428,
+      "grad_norm": 0.5507445485284483,
+      "learning_rate": 6.668827368845678e-05,
+      "loss": 0.5242,
+      "step": 552
+    },
+    {
+      "epoch": 1.7048169556840076,
+      "grad_norm": 0.6691206421402637,
+      "learning_rate": 6.662401189445176e-05,
+      "loss": 0.5244,
+      "step": 553
+    },
+    {
+      "epoch": 1.7078998073217726,
+      "grad_norm": 0.8901027618465831,
+      "learning_rate": 6.655962648918824e-05,
+      "loss": 0.5218,
+      "step": 554
+    },
+    {
+      "epoch": 1.7109826589595376,
+      "grad_norm": 1.0480343194938295,
+      "learning_rate": 6.649511777159791e-05,
+      "loss": 0.5275,
+      "step": 555
+    },
+    {
+      "epoch": 1.7140655105973024,
+      "grad_norm": 0.8558006886023909,
+      "learning_rate": 6.643048604118501e-05,
+      "loss": 0.5216,
+      "step": 556
+    },
+    {
+      "epoch": 1.7171483622350674,
+      "grad_norm": 0.6488056533116043,
+      "learning_rate": 6.636573159802487e-05,
+      "loss": 0.5211,
+      "step": 557
+    },
+    {
+      "epoch": 1.7202312138728324,
+      "grad_norm": 0.6890265088756017,
+      "learning_rate": 6.630085474276256e-05,
+      "loss": 0.5195,
+      "step": 558
+    },
+    {
+      "epoch": 1.7233140655105972,
+      "grad_norm": 0.805855619572633,
+      "learning_rate": 6.62358557766115e-05,
+      "loss": 0.5226,
+      "step": 559
+    },
+    {
+      "epoch": 1.7263969171483622,
+      "grad_norm": 0.6947900545942055,
+      "learning_rate": 6.617073500135208e-05,
+      "loss": 0.5209,
+      "step": 560
+    },
+    {
+      "epoch": 1.7294797687861272,
+      "grad_norm": 0.5353737643239515,
+      "learning_rate": 6.610549271933018e-05,
+      "loss": 0.5191,
+      "step": 561
+    },
+    {
+      "epoch": 1.732562620423892,
+      "grad_norm": 0.5192865394106787,
+      "learning_rate": 6.604012923345586e-05,
+      "loss": 0.5204,
+      "step": 562
+    },
+    {
+      "epoch": 1.735645472061657,
+      "grad_norm": 0.5570219815888906,
+      "learning_rate": 6.59746448472019e-05,
+      "loss": 0.5174,
+      "step": 563
+    },
+    {
+      "epoch": 1.738728323699422,
+      "grad_norm": 0.5589921490012696,
+      "learning_rate": 6.590903986460236e-05,
+      "loss": 0.5208,
+      "step": 564
+    },
+    {
+      "epoch": 1.7418111753371868,
+      "grad_norm": 0.5982247818753393,
+      "learning_rate": 6.584331459025128e-05,
+      "loss": 0.5214,
+      "step": 565
+    },
+    {
+      "epoch": 1.744894026974952,
+      "grad_norm": 0.4767492479006169,
+      "learning_rate": 6.577746932930118e-05,
+      "loss": 0.5205,
+      "step": 566
+    },
+    {
+      "epoch": 1.7479768786127168,
+      "grad_norm": 0.3837994174465851,
+      "learning_rate": 6.571150438746157e-05,
+      "loss": 0.5211,
+      "step": 567
+    },
+    {
+      "epoch": 1.7510597302504816,
+      "grad_norm": 0.5545618210428261,
+      "learning_rate": 6.564542007099776e-05,
+      "loss": 0.5198,
+      "step": 568
+    },
+    {
+      "epoch": 1.7541425818882468,
+      "grad_norm": 0.5950984317874429,
+      "learning_rate": 6.557921668672923e-05,
+      "loss": 0.5195,
+      "step": 569
+    },
+    {
+      "epoch": 1.7572254335260116,
+      "grad_norm": 0.5533190250490863,
+      "learning_rate": 6.551289454202824e-05,
+      "loss": 0.5194,
+      "step": 570
+    },
+    {
+      "epoch": 1.7603082851637764,
+      "grad_norm": 0.5850573143227845,
+      "learning_rate": 6.54464539448185e-05,
+      "loss": 0.5176,
+      "step": 571
+    },
+    {
+      "epoch": 1.7633911368015416,
+      "grad_norm": 0.5557904262230553,
+      "learning_rate": 6.537989520357362e-05,
+      "loss": 0.5187,
+      "step": 572
+    },
+    {
+      "epoch": 1.7664739884393064,
+      "grad_norm": 0.5107738867072339,
+      "learning_rate": 6.531321862731581e-05,
+      "loss": 0.5203,
+      "step": 573
+    },
+    {
+      "epoch": 1.7695568400770711,
+      "grad_norm": 0.5081535817028032,
+      "learning_rate": 6.524642452561427e-05,
+      "loss": 0.5183,
+      "step": 574
+    },
+    {
+      "epoch": 1.7726396917148364,
+      "grad_norm": 0.5631473100850715,
+      "learning_rate": 6.517951320858394e-05,
+      "loss": 0.5184,
+      "step": 575
+    },
+    {
+      "epoch": 1.7757225433526012,
+      "grad_norm": 0.7106308168223683,
+      "learning_rate": 6.511248498688396e-05,
+      "loss": 0.5222,
+      "step": 576
+    },
+    {
+      "epoch": 1.778805394990366,
+      "grad_norm": 0.8574456820987425,
+      "learning_rate": 6.504534017171615e-05,
+      "loss": 0.5184,
+      "step": 577
+    },
+    {
+      "epoch": 1.7818882466281312,
+      "grad_norm": 1.0162894475417772,
+      "learning_rate": 6.49780790748238e-05,
+      "loss": 0.5192,
+      "step": 578
+    },
+    {
+      "epoch": 1.784971098265896,
+      "grad_norm": 1.0824241759604394,
+      "learning_rate": 6.491070200848997e-05,
+      "loss": 0.5241,
+      "step": 579
+    },
+    {
+      "epoch": 1.7880539499036607,
+      "grad_norm": 0.7585758542530944,
+      "learning_rate": 6.484320928553618e-05,
+      "loss": 0.516,
+      "step": 580
+    },
+    {
+      "epoch": 1.791136801541426,
+      "grad_norm": 0.41210423366969035,
+      "learning_rate": 6.477560121932095e-05,
+      "loss": 0.5203,
+      "step": 581
+    },
+    {
+      "epoch": 1.7942196531791907,
+      "grad_norm": 0.35834639114914346,
+      "learning_rate": 6.470787812373829e-05,
+      "loss": 0.5198,
+      "step": 582
+    },
+    {
+      "epoch": 1.7973025048169555,
+      "grad_norm": 0.5511680308197465,
+      "learning_rate": 6.464004031321627e-05,
+      "loss": 0.5204,
+      "step": 583
+    },
+    {
+      "epoch": 1.8003853564547208,
+      "grad_norm": 0.646290355388511,
+      "learning_rate": 6.457208810271562e-05,
+      "loss": 0.5174,
+      "step": 584
+    },
+    {
+      "epoch": 1.8034682080924855,
+      "grad_norm": 0.6224811696106277,
+      "learning_rate": 6.450402180772811e-05,
+      "loss": 0.5174,
+      "step": 585
+    },
+    {
+      "epoch": 1.8065510597302503,
+      "grad_norm": 0.5719675586177476,
+      "learning_rate": 6.443584174427532e-05,
+      "loss": 0.5178,
+      "step": 586
+    },
+    {
+      "epoch": 1.8096339113680155,
+      "grad_norm": 0.501931526928349,
+      "learning_rate": 6.43675482289069e-05,
+      "loss": 0.5195,
+      "step": 587
+    },
+    {
+      "epoch": 1.8127167630057803,
+      "grad_norm": 0.47929312516512507,
+      "learning_rate": 6.429914157869934e-05,
+      "loss": 0.5214,
+      "step": 588
+    },
+    {
+      "epoch": 1.8157996146435453,
+      "grad_norm": 0.5986931171296265,
+      "learning_rate": 6.423062211125438e-05,
+      "loss": 0.5189,
+      "step": 589
+    },
+    {
+      "epoch": 1.8188824662813103,
+      "grad_norm": 0.6989905339669678,
+      "learning_rate": 6.416199014469751e-05,
+      "loss": 0.5209,
+      "step": 590
+    },
+    {
+      "epoch": 1.8219653179190751,
+      "grad_norm": 0.6698149747034,
+      "learning_rate": 6.40932459976766e-05,
+      "loss": 0.5167,
+      "step": 591
+    },
+    {
+      "epoch": 1.8250481695568401,
+      "grad_norm": 0.5600269208075253,
+      "learning_rate": 6.402438998936027e-05,
+      "loss": 0.517,
+      "step": 592
+    },
+    {
+      "epoch": 1.8281310211946051,
+      "grad_norm": 0.5551553070549329,
+      "learning_rate": 6.395542243943659e-05,
+      "loss": 0.5164,
+      "step": 593
+    },
+    {
+      "epoch": 1.83121387283237,
+      "grad_norm": 0.5997582030139671,
+      "learning_rate": 6.388634366811146e-05,
+      "loss": 0.516,
+      "step": 594
+    },
+    {
+      "epoch": 1.834296724470135,
+      "grad_norm": 0.51174538976797,
+      "learning_rate": 6.381715399610713e-05,
+      "loss": 0.5199,
+      "step": 595
+    },
+    {
+      "epoch": 1.8373795761079,
+      "grad_norm": 0.537804629316173,
+      "learning_rate": 6.374785374466081e-05,
+      "loss": 0.5179,
+      "step": 596
+    },
+    {
+      "epoch": 1.8404624277456647,
+      "grad_norm": 0.6261001711674018,
+      "learning_rate": 6.367844323552304e-05,
+      "loss": 0.5226,
+      "step": 597
+    },
+    {
+      "epoch": 1.8435452793834297,
+      "grad_norm": 0.5666878983357223,
+      "learning_rate": 6.360892279095636e-05,
+      "loss": 0.518,
+      "step": 598
+    },
+    {
+      "epoch": 1.8466281310211947,
+      "grad_norm": 0.6001363957417697,
+      "learning_rate": 6.353929273373365e-05,
+      "loss": 0.5147,
+      "step": 599
+    },
+    {
+      "epoch": 1.8497109826589595,
+      "grad_norm": 0.6519164164863905,
+      "learning_rate": 6.346955338713672e-05,
+      "loss": 0.513,
+      "step": 600
+    },
+    {
+      "epoch": 1.8527938342967245,
+      "grad_norm": 0.7505299752679043,
+      "learning_rate": 6.339970507495482e-05,
+      "loss": 0.5168,
+      "step": 601
+    },
+    {
+      "epoch": 1.8558766859344895,
+      "grad_norm": 0.9868226038839831,
+      "learning_rate": 6.332974812148309e-05,
+      "loss": 0.5179,
+      "step": 602
+    },
+    {
+      "epoch": 1.8589595375722543,
+      "grad_norm": 1.1655157991583023,
+      "learning_rate": 6.325968285152107e-05,
+      "loss": 0.521,
+      "step": 603
+    },
+    {
+      "epoch": 1.8620423892100193,
+      "grad_norm": 0.6062141569215225,
+      "learning_rate": 6.318950959037122e-05,
+      "loss": 0.5187,
+      "step": 604
+    },
+    {
+      "epoch": 1.8651252408477843,
+      "grad_norm": 0.4286228861054574,
+      "learning_rate": 6.311922866383737e-05,
+      "loss": 0.5132,
+      "step": 605
+    },
+    {
+      "epoch": 1.868208092485549,
+      "grad_norm": 0.6401620967698752,
+      "learning_rate": 6.304884039822322e-05,
+      "loss": 0.5168,
+      "step": 606
+    },
+    {
+      "epoch": 1.871290944123314,
+      "grad_norm": 0.897466051101545,
+      "learning_rate": 6.297834512033082e-05,
+      "loss": 0.5147,
+      "step": 607
+    },
+    {
+      "epoch": 1.874373795761079,
+      "grad_norm": 1.1341225652459093,
+      "learning_rate": 6.290774315745913e-05,
+      "loss": 0.5166,
+      "step": 608
+    },
+    {
+      "epoch": 1.8774566473988439,
+      "grad_norm": 0.6252528374698536,
+      "learning_rate": 6.283703483740234e-05,
+      "loss": 0.516,
+      "step": 609
+    },
+    {
+      "epoch": 1.8805394990366089,
+      "grad_norm": 0.33012275419133064,
+      "learning_rate": 6.276622048844848e-05,
+      "loss": 0.5206,
+      "step": 610
+    },
+    {
+      "epoch": 1.883622350674374,
+      "grad_norm": 0.36542415004897355,
+      "learning_rate": 6.269530043937787e-05,
+      "loss": 0.5161,
+      "step": 611
+    },
+    {
+      "epoch": 1.8867052023121387,
+      "grad_norm": 0.4320966176976742,
+      "learning_rate": 6.262427501946155e-05,
+      "loss": 0.519,
+      "step": 612
+    },
+    {
+      "epoch": 1.8897880539499037,
+      "grad_norm": 0.47714172777347147,
+      "learning_rate": 6.255314455845982e-05,
+      "loss": 0.5184,
+      "step": 613
+    },
+    {
+      "epoch": 1.8928709055876687,
+      "grad_norm": 0.5949024452833247,
+      "learning_rate": 6.248190938662064e-05,
+      "loss": 0.5128,
+      "step": 614
+    },
+    {
+      "epoch": 1.8959537572254335,
+      "grad_norm": 0.6892383825231114,
+      "learning_rate": 6.241056983467811e-05,
+      "loss": 0.5188,
+      "step": 615
+    },
+    {
+      "epoch": 1.8990366088631985,
+      "grad_norm": 0.6752101897438981,
+      "learning_rate": 6.2339126233851e-05,
+      "loss": 0.517,
+      "step": 616
+    },
+    {
+      "epoch": 1.9021194605009635,
+      "grad_norm": 0.6464474782435111,
+      "learning_rate": 6.226757891584113e-05,
+      "loss": 0.5158,
+      "step": 617
+    },
+    {
+      "epoch": 1.9052023121387283,
+      "grad_norm": 0.6286668345792895,
+      "learning_rate": 6.219592821283188e-05,
+      "loss": 0.517,
+      "step": 618
+    },
+    {
+      "epoch": 1.9082851637764933,
+      "grad_norm": 0.602028863571272,
+      "learning_rate": 6.212417445748661e-05,
+      "loss": 0.5155,
+      "step": 619
+    },
+    {
+      "epoch": 1.9113680154142583,
+      "grad_norm": 0.6713446849303493,
+      "learning_rate": 6.205231798294715e-05,
+      "loss": 0.5155,
+      "step": 620
+    },
+    {
+      "epoch": 1.914450867052023,
+      "grad_norm": 0.7997873761382751,
+      "learning_rate": 6.198035912283225e-05,
+      "loss": 0.5125,
+      "step": 621
+    },
+    {
+      "epoch": 1.917533718689788,
+      "grad_norm": 0.8629859425248647,
+      "learning_rate": 6.190829821123598e-05,
+      "loss": 0.5169,
+      "step": 622
+    },
+    {
+      "epoch": 1.920616570327553,
+      "grad_norm": 0.8845256946441112,
+      "learning_rate": 6.183613558272629e-05,
+      "loss": 0.5185,
+      "step": 623
+    },
+    {
+      "epoch": 1.9236994219653178,
+      "grad_norm": 0.8191864842462326,
+      "learning_rate": 6.176387157234331e-05,
+      "loss": 0.5114,
+      "step": 624
+    },
+    {
+      "epoch": 1.9267822736030829,
+      "grad_norm": 0.7208490060199927,
+      "learning_rate": 6.169150651559793e-05,
+      "loss": 0.518,
+      "step": 625
+    },
+    {
+      "epoch": 1.9298651252408479,
+      "grad_norm": 0.5942925297899428,
+      "learning_rate": 6.161904074847013e-05,
+      "loss": 0.5148,
+      "step": 626
+    },
+    {
+      "epoch": 1.9329479768786126,
+      "grad_norm": 0.5059212468450976,
+      "learning_rate": 6.154647460740752e-05,
+      "loss": 0.5182,
+      "step": 627
+    },
+    {
+      "epoch": 1.9360308285163776,
+      "grad_norm": 0.4760404863736317,
+      "learning_rate": 6.14738084293237e-05,
+      "loss": 0.5166,
+      "step": 628
+    },
+    {
+      "epoch": 1.9391136801541426,
+      "grad_norm": 0.4783801757341339,
+      "learning_rate": 6.140104255159677e-05,
+      "loss": 0.5147,
+      "step": 629
+    },
+    {
+      "epoch": 1.9421965317919074,
+      "grad_norm": 0.45805457463610433,
+      "learning_rate": 6.132817731206766e-05,
+      "loss": 0.5187,
+      "step": 630
+    },
+    {
+      "epoch": 1.9452793834296724,
+      "grad_norm": 0.44580433463428015,
+      "learning_rate": 6.125521304903865e-05,
+      "loss": 0.5181,
+      "step": 631
+    },
+    {
+      "epoch": 1.9483622350674374,
+      "grad_norm": 0.4527138693897574,
+      "learning_rate": 6.118215010127178e-05,
+      "loss": 0.518,
+      "step": 632
+    },
+    {
+      "epoch": 1.9514450867052022,
+      "grad_norm": 0.4459297503587521,
+      "learning_rate": 6.110898880798727e-05,
+      "loss": 0.5158,
+      "step": 633
+    },
+    {
+      "epoch": 1.9545279383429672,
+      "grad_norm": 0.350228449691934,
+      "learning_rate": 6.103572950886193e-05,
+      "loss": 0.5171,
+      "step": 634
+    },
+    {
+      "epoch": 1.9576107899807322,
+      "grad_norm": 0.3292117544319114,
+      "learning_rate": 6.096237254402759e-05,
+      "loss": 0.517,
+      "step": 635
+    },
+    {
+      "epoch": 1.960693641618497,
+      "grad_norm": 0.38671524281976855,
+      "learning_rate": 6.088891825406954e-05,
+      "loss": 0.5114,
+      "step": 636
+    },
+    {
+      "epoch": 1.963776493256262,
+      "grad_norm": 0.5112076563241872,
+      "learning_rate": 6.0815366980024954e-05,
+      "loss": 0.5133,
+      "step": 637
+    },
+    {
+      "epoch": 1.966859344894027,
+      "grad_norm": 0.5613595085849983,
+      "learning_rate": 6.074171906338123e-05,
+      "loss": 0.5206,
+      "step": 638
+    },
+    {
+      "epoch": 1.9699421965317918,
+      "grad_norm": 0.5054757411416915,
+      "learning_rate": 6.0667974846074524e-05,
+      "loss": 0.5153,
+      "step": 639
+    },
+    {
+      "epoch": 1.9730250481695568,
+      "grad_norm": 0.42413464981122034,
+      "learning_rate": 6.0594134670488075e-05,
+      "loss": 0.5148,
+      "step": 640
+    },
+    {
+      "epoch": 1.9761078998073218,
+      "grad_norm": 0.3864595705801421,
+      "learning_rate": 6.0520198879450645e-05,
+      "loss": 0.5112,
+      "step": 641
+    },
+    {
+      "epoch": 1.9791907514450866,
+      "grad_norm": 0.4259508271073006,
+      "learning_rate": 6.044616781623493e-05,
+      "loss": 0.5105,
+      "step": 642
+    },
+    {
+      "epoch": 1.9822736030828516,
+      "grad_norm": 0.5428399058734784,
+      "learning_rate": 6.037204182455595e-05,
+      "loss": 0.5154,
+      "step": 643
+    },
+    {
+      "epoch": 1.9853564547206166,
+      "grad_norm": 0.6513074290234719,
+      "learning_rate": 6.029782124856945e-05,
+      "loss": 0.5152,
+      "step": 644
+    },
+    {
+      "epoch": 1.9884393063583814,
+      "grad_norm": 0.7150868407837347,
+      "learning_rate": 6.022350643287036e-05,
+      "loss": 0.5147,
+      "step": 645
+    },
+    {
+      "epoch": 1.9915221579961464,
+      "grad_norm": 0.7154923878240175,
+      "learning_rate": 6.0149097722491106e-05,
+      "loss": 0.5162,
+      "step": 646
+    },
+    {
+      "epoch": 1.9946050096339114,
+      "grad_norm": 0.7383726044093668,
+      "learning_rate": 6.0074595462900086e-05,
+      "loss": 0.5145,
+      "step": 647
+    },
+    {
+      "epoch": 1.9976878612716762,
+      "grad_norm": 0.8211133768247235,
+      "learning_rate": 6.000000000000001e-05,
+      "loss": 0.517,
+      "step": 648
+    },
+    {
+      "epoch": 2.0007707129094414,
+      "grad_norm": 1.4066264815483163,
+      "learning_rate": 5.9925311680126315e-05,
+      "loss": 0.8288,
+      "step": 649
+    },
+    {
+      "epoch": 2.003853564547206,
+      "grad_norm": 1.2993671249853302,
+      "learning_rate": 5.985053085004558e-05,
+      "loss": 0.4912,
+      "step": 650
+    },
+    {
+      "epoch": 2.006936416184971,
+      "grad_norm": 0.4876086402676008,
+      "learning_rate": 5.977565785695388e-05,
+      "loss": 0.4873,
+      "step": 651
+    },
+    {
+      "epoch": 2.010019267822736,
+      "grad_norm": 0.5083675172550183,
+      "learning_rate": 5.97006930484752e-05,
+      "loss": 0.489,
+      "step": 652
+    },
+    {
+      "epoch": 2.013102119460501,
+      "grad_norm": 0.9314943305854423,
+      "learning_rate": 5.962563677265978e-05,
+      "loss": 0.4913,
+      "step": 653
+    },
+    {
+      "epoch": 2.0161849710982658,
+      "grad_norm": 1.0920602336437433,
+      "learning_rate": 5.955048937798257e-05,
+      "loss": 0.4911,
+      "step": 654
+    },
+    {
+      "epoch": 2.019267822736031,
+      "grad_norm": 0.5348185666920724,
+      "learning_rate": 5.947525121334156e-05,
+      "loss": 0.4873,
+      "step": 655
+    },
+    {
+      "epoch": 2.022350674373796,
+      "grad_norm": 0.4840172624697722,
+      "learning_rate": 5.9399922628056133e-05,
+      "loss": 0.4867,
+      "step": 656
+    },
+    {
+      "epoch": 2.0254335260115606,
+      "grad_norm": 0.7666931989938438,
+      "learning_rate": 5.9324503971865545e-05,
+      "loss": 0.4887,
+      "step": 657
+    },
+    {
+      "epoch": 2.028516377649326,
+      "grad_norm": 0.8552612231943164,
+      "learning_rate": 5.9248995594927185e-05,
+      "loss": 0.4891,
+      "step": 658
+    },
+    {
+      "epoch": 2.0315992292870906,
+      "grad_norm": 0.731821313909628,
+      "learning_rate": 5.917339784781502e-05,
+      "loss": 0.4916,
+      "step": 659
+    },
+    {
+      "epoch": 2.0346820809248554,
+      "grad_norm": 0.4980210774834564,
+      "learning_rate": 5.909771108151797e-05,
+      "loss": 0.4908,
+      "step": 660
+    },
+    {
+      "epoch": 2.0377649325626206,
+      "grad_norm": 0.487975674421446,
+      "learning_rate": 5.9021935647438206e-05,
+      "loss": 0.4924,
+      "step": 661
+    },
+    {
+      "epoch": 2.0408477842003854,
+      "grad_norm": 0.5642901422556802,
+      "learning_rate": 5.8946071897389634e-05,
+      "loss": 0.4903,
+      "step": 662
+    },
+    {
+      "epoch": 2.04393063583815,
+      "grad_norm": 0.5374195233203573,
+      "learning_rate": 5.887012018359615e-05,
+      "loss": 0.4927,
+      "step": 663
+    },
+    {
+      "epoch": 2.0470134874759154,
+      "grad_norm": 0.4871174485783199,
+      "learning_rate": 5.8794080858690064e-05,
+      "loss": 0.4883,
+      "step": 664
+    },
+    {
+      "epoch": 2.05009633911368,
+      "grad_norm": 0.5117880454244503,
+      "learning_rate": 5.871795427571046e-05,
+      "loss": 0.4884,
+      "step": 665
+    },
+    {
+      "epoch": 2.053179190751445,
+      "grad_norm": 0.5486779244692324,
+      "learning_rate": 5.8641740788101566e-05,
+      "loss": 0.4899,
+      "step": 666
+    },
+    {
+      "epoch": 2.05626204238921,
+      "grad_norm": 0.5370112426612905,
+      "learning_rate": 5.856544074971103e-05,
+      "loss": 0.4877,
+      "step": 667
+    },
+    {
+      "epoch": 2.059344894026975,
+      "grad_norm": 0.482073582631281,
+      "learning_rate": 5.848905451478841e-05,
+      "loss": 0.4887,
+      "step": 668
+    },
+    {
+      "epoch": 2.0624277456647397,
+      "grad_norm": 0.38722549529614364,
+      "learning_rate": 5.841258243798344e-05,
+      "loss": 0.4851,
+      "step": 669
+    },
+    {
+      "epoch": 2.065510597302505,
+      "grad_norm": 0.4055525959046836,
+      "learning_rate": 5.8336024874344384e-05,
+      "loss": 0.4887,
+      "step": 670
+    },
+    {
+      "epoch": 2.0685934489402698,
+      "grad_norm": 0.5422362860634989,
+      "learning_rate": 5.825938217931644e-05,
+      "loss": 0.4883,
+      "step": 671
+    },
+    {
+      "epoch": 2.0716763005780345,
+      "grad_norm": 0.5116562010719634,
+      "learning_rate": 5.818265470874003e-05,
+      "loss": 0.4916,
+      "step": 672
+    },
+    {
+      "epoch": 2.0747591522157998,
+      "grad_norm": 0.4063080271272241,
+      "learning_rate": 5.810584281884921e-05,
+      "loss": 0.4907,
+      "step": 673
+    },
+    {
+      "epoch": 2.0778420038535645,
+      "grad_norm": 0.504070060061359,
+      "learning_rate": 5.802894686626994e-05,
+      "loss": 0.4911,
+      "step": 674
+    },
+    {
+      "epoch": 2.0809248554913293,
+      "grad_norm": 0.6120802048085145,
+      "learning_rate": 5.79519672080185e-05,
+      "loss": 0.489,
+      "step": 675
+    },
+    {
+      "epoch": 2.0840077071290946,
+      "grad_norm": 0.6618100915328166,
+      "learning_rate": 5.7874904201499795e-05,
+      "loss": 0.4891,
+      "step": 676
+    },
+    {
+      "epoch": 2.0870905587668593,
+      "grad_norm": 0.5978129184480812,
+      "learning_rate": 5.77977582045057e-05,
+      "loss": 0.4903,
+      "step": 677
+    },
+    {
+      "epoch": 2.090173410404624,
+      "grad_norm": 0.47879977496946,
+      "learning_rate": 5.772052957521341e-05,
+      "loss": 0.4883,
+      "step": 678
+    },
+    {
+      "epoch": 2.0932562620423893,
+      "grad_norm": 0.548955837794539,
+      "learning_rate": 5.764321867218377e-05,
+      "loss": 0.4911,
+      "step": 679
+    },
+    {
+      "epoch": 2.096339113680154,
+      "grad_norm": 0.5934072563796927,
+      "learning_rate": 5.75658258543596e-05,
+      "loss": 0.486,
+      "step": 680
+    },
+    {
+      "epoch": 2.099421965317919,
+      "grad_norm": 0.4834822026331229,
+      "learning_rate": 5.748835148106404e-05,
+      "loss": 0.4857,
+      "step": 681
+    },
+    {
+      "epoch": 2.102504816955684,
+      "grad_norm": 0.4491809460985718,
+      "learning_rate": 5.741079591199887e-05,
+      "loss": 0.4911,
+      "step": 682
+    },
+    {
+      "epoch": 2.105587668593449,
+      "grad_norm": 0.5259397218238746,
+      "learning_rate": 5.7333159507242885e-05,
+      "loss": 0.4869,
+      "step": 683
+    },
+    {
+      "epoch": 2.1086705202312137,
+      "grad_norm": 0.5319060030589742,
+      "learning_rate": 5.7255442627250146e-05,
+      "loss": 0.4879,
+      "step": 684
+    },
+    {
+      "epoch": 2.111753371868979,
+      "grad_norm": 0.4615856856757023,
+      "learning_rate": 5.717764563284837e-05,
+      "loss": 0.4862,
+      "step": 685
+    },
+    {
+      "epoch": 2.1148362235067437,
+      "grad_norm": 0.434911452208656,
+      "learning_rate": 5.709976888523725e-05,
+      "loss": 0.4867,
+      "step": 686
+    },
+    {
+      "epoch": 2.1179190751445085,
+      "grad_norm": 0.4313500082323918,
+      "learning_rate": 5.70218127459867e-05,
+      "loss": 0.4888,
+      "step": 687
+    },
+    {
+      "epoch": 2.1210019267822737,
+      "grad_norm": 0.5327742733371285,
+      "learning_rate": 5.694377757703532e-05,
+      "loss": 0.4948,
+      "step": 688
+    },
+    {
+      "epoch": 2.1240847784200385,
+      "grad_norm": 0.561957717262307,
+      "learning_rate": 5.686566374068857e-05,
+      "loss": 0.4892,
+      "step": 689
+    },
+    {
+      "epoch": 2.1271676300578033,
+      "grad_norm": 0.45353761812655563,
+      "learning_rate": 5.678747159961717e-05,
+      "loss": 0.4929,
+      "step": 690
+    },
+    {
+      "epoch": 2.1302504816955685,
+      "grad_norm": 0.32811248715853875,
+      "learning_rate": 5.670920151685541e-05,
+      "loss": 0.4837,
+      "step": 691
+    },
+    {
+      "epoch": 2.1333333333333333,
+      "grad_norm": 0.3963542284941922,
+      "learning_rate": 5.663085385579944e-05,
+      "loss": 0.4906,
+      "step": 692
+    },
+    {
+      "epoch": 2.136416184971098,
+      "grad_norm": 0.5132326122912824,
+      "learning_rate": 5.6552428980205575e-05,
+      "loss": 0.4892,
+      "step": 693
+    },
+    {
+      "epoch": 2.1394990366088633,
+      "grad_norm": 0.5352078356708485,
+      "learning_rate": 5.647392725418867e-05,
+      "loss": 0.4868,
+      "step": 694
+    },
+    {
+      "epoch": 2.142581888246628,
+      "grad_norm": 0.4791720907442877,
+      "learning_rate": 5.639534904222034e-05,
+      "loss": 0.489,
+      "step": 695
+    },
+    {
+      "epoch": 2.145664739884393,
+      "grad_norm": 0.4411890616528477,
+      "learning_rate": 5.631669470912734e-05,
+      "loss": 0.4893,
+      "step": 696
+    },
+    {
+      "epoch": 2.148747591522158,
+      "grad_norm": 0.35678825428033556,
+      "learning_rate": 5.623796462008986e-05,
+      "loss": 0.4887,
+      "step": 697
+    },
+    {
+      "epoch": 2.151830443159923,
+      "grad_norm": 0.41370413297797065,
+      "learning_rate": 5.6159159140639755e-05,
+      "loss": 0.4846,
+      "step": 698
+    },
+    {
+      "epoch": 2.1549132947976877,
+      "grad_norm": 0.4451224867509384,
+      "learning_rate": 5.6080278636658964e-05,
+      "loss": 0.4901,
+      "step": 699
+    },
+    {
+      "epoch": 2.157996146435453,
+      "grad_norm": 0.3995290453308094,
+      "learning_rate": 5.600132347437771e-05,
+      "loss": 0.4869,
+      "step": 700
+    },
+    {
+      "epoch": 2.1610789980732177,
+      "grad_norm": 0.48730989419680965,
+      "learning_rate": 5.592229402037289e-05,
+      "loss": 0.4918,
+      "step": 701
+    },
+    {
+      "epoch": 2.1641618497109825,
+      "grad_norm": 0.4861734232038218,
+      "learning_rate": 5.584319064156628e-05,
+      "loss": 0.4901,
+      "step": 702
+    },
+    {
+      "epoch": 2.1672447013487477,
+      "grad_norm": 0.48434949941957095,
+      "learning_rate": 5.57640137052229e-05,
+      "loss": 0.4866,
+      "step": 703
+    },
+    {
+      "epoch": 2.1703275529865125,
+      "grad_norm": 0.4784461883302772,
+      "learning_rate": 5.5684763578949294e-05,
+      "loss": 0.4822,
+      "step": 704
+    },
+    {
+      "epoch": 2.1734104046242773,
+      "grad_norm": 0.35406282685324536,
+      "learning_rate": 5.56054406306918e-05,
+      "loss": 0.488,
+      "step": 705
+    },
+    {
+      "epoch": 2.1764932562620425,
+      "grad_norm": 0.4591547184504055,
+      "learning_rate": 5.552604522873487e-05,
+      "loss": 0.4917,
+      "step": 706
+    },
+    {
+      "epoch": 2.1795761078998073,
+      "grad_norm": 0.5028749183562098,
+      "learning_rate": 5.544657774169934e-05,
+      "loss": 0.4896,
+      "step": 707
+    },
+    {
+      "epoch": 2.182658959537572,
+      "grad_norm": 0.38082017635098087,
+      "learning_rate": 5.5367038538540706e-05,
+      "loss": 0.4886,
+      "step": 708
+    },
+    {
+      "epoch": 2.1857418111753373,
+      "grad_norm": 0.4513361898787459,
+      "learning_rate": 5.528742798854748e-05,
+      "loss": 0.486,
+      "step": 709
+    },
+    {
+      "epoch": 2.188824662813102,
+      "grad_norm": 0.46096918067422726,
+      "learning_rate": 5.520774646133939e-05,
+      "loss": 0.4884,
+      "step": 710
+    },
+    {
+      "epoch": 2.191907514450867,
+      "grad_norm": 0.46447763398893116,
+      "learning_rate": 5.5127994326865706e-05,
+      "loss": 0.4908,
+      "step": 711
+    },
+    {
+      "epoch": 2.194990366088632,
+      "grad_norm": 0.3997019664887476,
+      "learning_rate": 5.504817195540352e-05,
+      "loss": 0.4895,
+      "step": 712
+    },
+    {
+      "epoch": 2.198073217726397,
+      "grad_norm": 0.3559478755532233,
+      "learning_rate": 5.496827971755603e-05,
+      "loss": 0.4866,
+      "step": 713
+    },
+    {
+      "epoch": 2.2011560693641616,
+      "grad_norm": 0.524842143489639,
+      "learning_rate": 5.488831798425081e-05,
+      "loss": 0.488,
+      "step": 714
+    },
+    {
+      "epoch": 2.204238921001927,
+      "grad_norm": 0.5849548759476255,
+      "learning_rate": 5.480828712673808e-05,
+      "loss": 0.4917,
+      "step": 715
+    },
+    {
+      "epoch": 2.2073217726396916,
+      "grad_norm": 0.4961079680988736,
+      "learning_rate": 5.472818751658898e-05,
+      "loss": 0.4888,
+      "step": 716
+    },
+    {
+      "epoch": 2.210404624277457,
+      "grad_norm": 0.3836907969118893,
+      "learning_rate": 5.4648019525693924e-05,
+      "loss": 0.4931,
+      "step": 717
+    },
+    {
+      "epoch": 2.2134874759152217,
+      "grad_norm": 0.33033646419564455,
+      "learning_rate": 5.456778352626073e-05,
+      "loss": 0.4873,
+      "step": 718
+    },
+    {
+      "epoch": 2.2165703275529864,
+      "grad_norm": 0.35034463512347164,
+      "learning_rate": 5.448747989081303e-05,
+      "loss": 0.4904,
+      "step": 719
+    },
+    {
+      "epoch": 2.2196531791907512,
+      "grad_norm": 0.44827616095651585,
+      "learning_rate": 5.440710899218842e-05,
+      "loss": 0.4884,
+      "step": 720
+    },
+    {
+      "epoch": 2.2227360308285165,
+      "grad_norm": 0.3923115469752325,
+      "learning_rate": 5.4326671203536844e-05,
+      "loss": 0.49,
+      "step": 721
+    },
+    {
+      "epoch": 2.2258188824662812,
+      "grad_norm": 0.3498819615622833,
+      "learning_rate": 5.424616689831877e-05,
+      "loss": 0.4881,
+      "step": 722
+    },
+    {
+      "epoch": 2.2289017341040465,
+      "grad_norm": 0.38175241870027243,
+      "learning_rate": 5.416559645030351e-05,
+      "loss": 0.4916,
+      "step": 723
+    },
+    {
+      "epoch": 2.2319845857418112,
+      "grad_norm": 0.3266682367888484,
+      "learning_rate": 5.408496023356746e-05,
+      "loss": 0.4897,
+      "step": 724
+    },
+    {
+      "epoch": 2.235067437379576,
+      "grad_norm": 0.4129597690969217,
+      "learning_rate": 5.400425862249238e-05,
+      "loss": 0.4894,
+      "step": 725
+    },
+    {
+      "epoch": 2.238150289017341,
+      "grad_norm": 0.4641206974067704,
+      "learning_rate": 5.3923491991763626e-05,
+      "loss": 0.4903,
+      "step": 726
+    },
+    {
+      "epoch": 2.241233140655106,
+      "grad_norm": 0.4176405774941974,
+      "learning_rate": 5.3842660716368455e-05,
+      "loss": 0.4901,
+      "step": 727
+    },
+    {
+      "epoch": 2.244315992292871,
+      "grad_norm": 0.45684072775303247,
+      "learning_rate": 5.376176517159424e-05,
+      "loss": 0.4856,
+      "step": 728
+    },
+    {
+      "epoch": 2.247398843930636,
+      "grad_norm": 0.5111984428674652,
+      "learning_rate": 5.368080573302676e-05,
+      "loss": 0.4892,
+      "step": 729
+    },
+    {
+      "epoch": 2.250481695568401,
+      "grad_norm": 0.4901448849519485,
+      "learning_rate": 5.3599782776548434e-05,
+      "loss": 0.4886,
+      "step": 730
+    },
+    {
+      "epoch": 2.2535645472061656,
+      "grad_norm": 0.37902270891007506,
+      "learning_rate": 5.351869667833658e-05,
+      "loss": 0.4902,
+      "step": 731
+    },
+    {
+      "epoch": 2.2566473988439304,
+      "grad_norm": 0.26179156495528033,
+      "learning_rate": 5.3437547814861694e-05,
+      "loss": 0.4821,
+      "step": 732
+    },
+    {
+      "epoch": 2.2597302504816956,
+      "grad_norm": 0.3106615902146888,
+      "learning_rate": 5.335633656288567e-05,
+      "loss": 0.4903,
+      "step": 733
+    },
+    {
+      "epoch": 2.2628131021194604,
+      "grad_norm": 0.49659528708725487,
+      "learning_rate": 5.327506329946005e-05,
+      "loss": 0.4846,
+      "step": 734
+    },
+    {
+      "epoch": 2.2658959537572256,
+      "grad_norm": 0.47755025245134997,
+      "learning_rate": 5.319372840192431e-05,
+      "loss": 0.4877,
+      "step": 735
+    },
+    {
+      "epoch": 2.2689788053949904,
+      "grad_norm": 0.41768160318115516,
+      "learning_rate": 5.311233224790407e-05,
+      "loss": 0.4928,
+      "step": 736
+    },
+    {
+      "epoch": 2.272061657032755,
+      "grad_norm": 0.38016666522190384,
+      "learning_rate": 5.303087521530936e-05,
+      "loss": 0.4911,
+      "step": 737
+    },
+    {
+      "epoch": 2.27514450867052,
+      "grad_norm": 0.35248137331259183,
+      "learning_rate": 5.294935768233285e-05,
+      "loss": 0.4854,
+      "step": 738
+    },
+    {
+      "epoch": 2.278227360308285,
+      "grad_norm": 0.372608533191928,
+      "learning_rate": 5.2867780027448104e-05,
+      "loss": 0.4871,
+      "step": 739
+    },
+    {
+      "epoch": 2.28131021194605,
+      "grad_norm": 0.3743868539471539,
+      "learning_rate": 5.2786142629407836e-05,
+      "loss": 0.4864,
+      "step": 740
+    },
+    {
+      "epoch": 2.284393063583815,
+      "grad_norm": 0.333928698169537,
+      "learning_rate": 5.270444586724213e-05,
+      "loss": 0.488,
+      "step": 741
+    },
+    {
+      "epoch": 2.28747591522158,
+      "grad_norm": 0.26548838198393543,
+      "learning_rate": 5.26226901202567e-05,
+      "loss": 0.4869,
+      "step": 742
+    },
+    {
+      "epoch": 2.290558766859345,
+      "grad_norm": 0.23852936335689645,
+      "learning_rate": 5.254087576803111e-05,
+      "loss": 0.4923,
+      "step": 743
+    },
+    {
+      "epoch": 2.29364161849711,
+      "grad_norm": 0.3344479155077712,
+      "learning_rate": 5.245900319041699e-05,
+      "loss": 0.4929,
+      "step": 744
+    },
+    {
+      "epoch": 2.296724470134875,
+      "grad_norm": 0.33186752990478924,
+      "learning_rate": 5.237707276753636e-05,
+      "loss": 0.4894,
+      "step": 745
+    },
+    {
+      "epoch": 2.2998073217726396,
+      "grad_norm": 0.3596306744147825,
+      "learning_rate": 5.229508487977974e-05,
+      "loss": 0.4891,
+      "step": 746
+    },
+    {
+      "epoch": 2.302890173410405,
+      "grad_norm": 0.4452675602848974,
+      "learning_rate": 5.2213039907804535e-05,
+      "loss": 0.4884,
+      "step": 747
+    },
+    {
+      "epoch": 2.3059730250481696,
+      "grad_norm": 0.3936693770445251,
+      "learning_rate": 5.21309382325331e-05,
+      "loss": 0.4907,
+      "step": 748
+    },
+    {
+      "epoch": 2.3090558766859344,
+      "grad_norm": 0.3477703079472612,
+      "learning_rate": 5.204878023515109e-05,
+      "loss": 0.4891,
+      "step": 749
+    },
+    {
+      "epoch": 2.3121387283236996,
+      "grad_norm": 0.31505068880534687,
+      "learning_rate": 5.1966566297105675e-05,
+      "loss": 0.4876,
+      "step": 750
+    },
+    {
+      "epoch": 2.3152215799614644,
+      "grad_norm": 0.2827157298540461,
+      "learning_rate": 5.188429680010371e-05,
+      "loss": 0.4815,
+      "step": 751
+    },
+    {
+      "epoch": 2.318304431599229,
+      "grad_norm": 0.3120722410130788,
+      "learning_rate": 5.1801972126110026e-05,
+      "loss": 0.4901,
+      "step": 752
+    },
+    {
+      "epoch": 2.3213872832369944,
+      "grad_norm": 0.3065080379074147,
+      "learning_rate": 5.171959265734563e-05,
+      "loss": 0.4863,
+      "step": 753
+    },
+    {
+      "epoch": 2.324470134874759,
+      "grad_norm": 0.34297868607872267,
+      "learning_rate": 5.163715877628593e-05,
+      "loss": 0.4868,
+      "step": 754
+    },
+    {
+      "epoch": 2.327552986512524,
+      "grad_norm": 0.48868065310356096,
+      "learning_rate": 5.155467086565896e-05,
+      "loss": 0.4881,
+      "step": 755
+    },
+    {
+      "epoch": 2.330635838150289,
+      "grad_norm": 0.675798008343738,
+      "learning_rate": 5.1472129308443616e-05,
+      "loss": 0.4874,
+      "step": 756
+    },
+    {
+      "epoch": 2.333718689788054,
+      "grad_norm": 0.7291815658627852,
+      "learning_rate": 5.138953448786784e-05,
+      "loss": 0.4888,
+      "step": 757
+    },
+    {
+      "epoch": 2.3368015414258188,
+      "grad_norm": 0.7252777494676279,
+      "learning_rate": 5.13068867874069e-05,
+      "loss": 0.4885,
+      "step": 758
+    },
+    {
+      "epoch": 2.339884393063584,
+      "grad_norm": 0.731323579579776,
+      "learning_rate": 5.122418659078156e-05,
+      "loss": 0.4899,
+      "step": 759
+    },
+    {
+      "epoch": 2.3429672447013488,
+      "grad_norm": 0.7292714399397713,
+      "learning_rate": 5.114143428195631e-05,
+      "loss": 0.4883,
+      "step": 760
+    },
+    {
+      "epoch": 2.3460500963391135,
+      "grad_norm": 0.59481902118636,
+      "learning_rate": 5.105863024513762e-05,
+      "loss": 0.4874,
+      "step": 761
+    },
+    {
+      "epoch": 2.3491329479768788,
+      "grad_norm": 0.45125830005502604,
+      "learning_rate": 5.097577486477207e-05,
+      "loss": 0.4867,
+      "step": 762
+    },
+    {
+      "epoch": 2.3522157996146436,
+      "grad_norm": 0.32449611306507464,
+      "learning_rate": 5.0892868525544676e-05,
+      "loss": 0.4912,
+      "step": 763
+    },
+    {
+      "epoch": 2.3552986512524083,
+      "grad_norm": 0.26805067823729667,
+      "learning_rate": 5.080991161237702e-05,
+      "loss": 0.4852,
+      "step": 764
+    },
+    {
+      "epoch": 2.3583815028901736,
+      "grad_norm": 0.4478581779470081,
+      "learning_rate": 5.07269045104255e-05,
+      "loss": 0.4909,
+      "step": 765
+    },
+    {
+      "epoch": 2.3614643545279383,
+      "grad_norm": 0.5625990525552398,
+      "learning_rate": 5.064384760507954e-05,
+      "loss": 0.4901,
+      "step": 766
+    },
+    {
+      "epoch": 2.364547206165703,
+      "grad_norm": 0.45702169933421266,
+      "learning_rate": 5.056074128195975e-05,
+      "loss": 0.4926,
+      "step": 767
+    },
+    {
+      "epoch": 2.3676300578034684,
+      "grad_norm": 0.368683408094175,
+      "learning_rate": 5.047758592691624e-05,
+      "loss": 0.487,
+      "step": 768
+    },
+    {
+      "epoch": 2.370712909441233,
+      "grad_norm": 0.3902951691939277,
+      "learning_rate": 5.039438192602674e-05,
+      "loss": 0.4891,
+      "step": 769
+    },
+    {
+      "epoch": 2.373795761078998,
+      "grad_norm": 0.3327019002979215,
+      "learning_rate": 5.031112966559483e-05,
+      "loss": 0.489,
+      "step": 770
+    },
+    {
+      "epoch": 2.376878612716763,
+      "grad_norm": 0.2720934524654798,
+      "learning_rate": 5.022782953214815e-05,
+      "loss": 0.4892,
+      "step": 771
+    },
+    {
+      "epoch": 2.379961464354528,
+      "grad_norm": 0.2949118897425433,
+      "learning_rate": 5.014448191243662e-05,
+      "loss": 0.4889,
+      "step": 772
+    },
+    {
+      "epoch": 2.3830443159922927,
+      "grad_norm": 0.45652966418413,
+      "learning_rate": 5.006108719343064e-05,
+      "loss": 0.4825,
+      "step": 773
+    },
+    {
+      "epoch": 2.386127167630058,
+      "grad_norm": 0.668496227768798,
+      "learning_rate": 4.9977645762319255e-05,
+      "loss": 0.4904,
+      "step": 774
+    },
+    {
+      "epoch": 2.3892100192678227,
+      "grad_norm": 0.7135935261559059,
+      "learning_rate": 4.9894158006508404e-05,
+      "loss": 0.4876,
+      "step": 775
+    },
+    {
+      "epoch": 2.3922928709055875,
+      "grad_norm": 0.6168447479074836,
+      "learning_rate": 4.981062431361912e-05,
+      "loss": 0.4897,
+      "step": 776
+    },
+    {
+      "epoch": 2.3953757225433527,
+      "grad_norm": 0.5325995820808088,
+      "learning_rate": 4.9727045071485674e-05,
+      "loss": 0.4897,
+      "step": 777
+    },
+    {
+      "epoch": 2.3984585741811175,
+      "grad_norm": 0.4797831032639185,
+      "learning_rate": 4.964342066815386e-05,
+      "loss": 0.4867,
+      "step": 778
+    },
+    {
+      "epoch": 2.4015414258188823,
+      "grad_norm": 0.46378006486715245,
+      "learning_rate": 4.955975149187912e-05,
+      "loss": 0.4885,
+      "step": 779
+    },
+    {
+      "epoch": 2.4046242774566475,
+      "grad_norm": 0.4240906389427457,
+      "learning_rate": 4.9476037931124765e-05,
+      "loss": 0.4884,
+      "step": 780
+    },
+    {
+      "epoch": 2.4077071290944123,
+      "grad_norm": 0.3152189210639397,
+      "learning_rate": 4.939228037456023e-05,
+      "loss": 0.4874,
+      "step": 781
+    },
+    {
+      "epoch": 2.410789980732177,
+      "grad_norm": 0.350271795351524,
+      "learning_rate": 4.930847921105913e-05,
+      "loss": 0.4884,
+      "step": 782
+    },
+    {
+      "epoch": 2.4138728323699423,
+      "grad_norm": 0.5209927578616728,
+      "learning_rate": 4.922463482969761e-05,
+      "loss": 0.4899,
+      "step": 783
+    },
+    {
+      "epoch": 2.416955684007707,
+      "grad_norm": 0.6621497117794022,
+      "learning_rate": 4.9140747619752456e-05,
+      "loss": 0.4876,
+      "step": 784
+    },
+    {
+      "epoch": 2.420038535645472,
+      "grad_norm": 0.7248686421597361,
+      "learning_rate": 4.9056817970699264e-05,
+      "loss": 0.492,
+      "step": 785
+    },
+    {
+      "epoch": 2.423121387283237,
+      "grad_norm": 0.5435678032887852,
+      "learning_rate": 4.897284627221071e-05,
+      "loss": 0.4894,
+      "step": 786
+    },
+    {
+      "epoch": 2.426204238921002,
+      "grad_norm": 0.3292044248597818,
+      "learning_rate": 4.8888832914154683e-05,
+      "loss": 0.4867,
+      "step": 787
+    },
+    {
+      "epoch": 2.4292870905587667,
+      "grad_norm": 0.2764788421750659,
+      "learning_rate": 4.880477828659249e-05,
+      "loss": 0.4886,
+      "step": 788
+    },
+    {
+      "epoch": 2.432369942196532,
+      "grad_norm": 0.34537123484084,
+      "learning_rate": 4.872068277977704e-05,
+      "loss": 0.4861,
+      "step": 789
+    },
+    {
+      "epoch": 2.4354527938342967,
+      "grad_norm": 0.41578120448892464,
+      "learning_rate": 4.863654678415105e-05,
+      "loss": 0.4889,
+      "step": 790
+    },
+    {
+      "epoch": 2.4385356454720615,
+      "grad_norm": 0.49694339718267483,
+      "learning_rate": 4.855237069034523e-05,
+      "loss": 0.4873,
+      "step": 791
+    },
+    {
+      "epoch": 2.4416184971098267,
+      "grad_norm": 0.5098054608105518,
+      "learning_rate": 4.846815488917644e-05,
+      "loss": 0.4896,
+      "step": 792
+    },
+    {
+      "epoch": 2.4447013487475915,
+      "grad_norm": 0.3753200984027443,
+      "learning_rate": 4.838389977164588e-05,
+      "loss": 0.4881,
+      "step": 793
+    },
+    {
+      "epoch": 2.4477842003853563,
+      "grad_norm": 0.3697266763717259,
+      "learning_rate": 4.829960572893734e-05,
+      "loss": 0.4875,
+      "step": 794
+    },
+    {
+      "epoch": 2.4508670520231215,
+      "grad_norm": 0.37959534956289726,
+      "learning_rate": 4.821527315241528e-05,
+      "loss": 0.4891,
+      "step": 795
+    },
+    {
+      "epoch": 2.4539499036608863,
+      "grad_norm": 0.32195001860090394,
+      "learning_rate": 4.813090243362312e-05,
+      "loss": 0.4895,
+      "step": 796
+    },
+    {
+      "epoch": 2.457032755298651,
+      "grad_norm": 0.24646168680575073,
+      "learning_rate": 4.804649396428132e-05,
+      "loss": 0.4835,
+      "step": 797
+    },
+    {
+      "epoch": 2.4601156069364163,
+      "grad_norm": 0.20605065713022086,
+      "learning_rate": 4.7962048136285626e-05,
+      "loss": 0.489,
+      "step": 798
+    },
+    {
+      "epoch": 2.463198458574181,
+      "grad_norm": 0.2521770642830829,
+      "learning_rate": 4.7877565341705256e-05,
+      "loss": 0.4847,
+      "step": 799
+    },
+    {
+      "epoch": 2.4662813102119463,
+      "grad_norm": 0.3985378724309409,
+      "learning_rate": 4.7793045972781024e-05,
+      "loss": 0.4899,
+      "step": 800
+    },
+    {
+      "epoch": 2.469364161849711,
+      "grad_norm": 0.5248431548312416,
+      "learning_rate": 4.7708490421923596e-05,
+      "loss": 0.4909,
+      "step": 801
+    },
+    {
+      "epoch": 2.472447013487476,
+      "grad_norm": 0.47215498694592456,
+      "learning_rate": 4.762389908171158e-05,
+      "loss": 0.4884,
+      "step": 802
+    },
+    {
+      "epoch": 2.4755298651252406,
+      "grad_norm": 0.3627755102518326,
+      "learning_rate": 4.7539272344889764e-05,
+      "loss": 0.4909,
+      "step": 803
+    },
+    {
+      "epoch": 2.478612716763006,
+      "grad_norm": 0.3492175105767242,
+      "learning_rate": 4.74546106043673e-05,
+      "loss": 0.4898,
+      "step": 804
+    },
+    {
+      "epoch": 2.4816955684007707,
+      "grad_norm": 0.3295569929430368,
+      "learning_rate": 4.736991425321581e-05,
+      "loss": 0.4833,
+      "step": 805
+    },
+    {
+      "epoch": 2.484778420038536,
+      "grad_norm": 0.4620381366561225,
+      "learning_rate": 4.728518368466766e-05,
+      "loss": 0.4894,
+      "step": 806
+    },
+    {
+      "epoch": 2.4878612716763007,
+      "grad_norm": 0.49648483788349856,
+      "learning_rate": 4.720041929211406e-05,
+      "loss": 0.4899,
+      "step": 807
+    },
+    {
+      "epoch": 2.4909441233140655,
+      "grad_norm": 0.38592522597300005,
+      "learning_rate": 4.711562146910325e-05,
+      "loss": 0.4848,
+      "step": 808
+    },
+    {
+      "epoch": 2.4940269749518302,
+      "grad_norm": 0.2527563363488598,
+      "learning_rate": 4.703079060933868e-05,
+      "loss": 0.4878,
+      "step": 809
+    },
+    {
+      "epoch": 2.4971098265895955,
+      "grad_norm": 0.5197978839002917,
+      "learning_rate": 4.694592710667723e-05,
+      "loss": 0.4874,
+      "step": 810
+    },
+    {
+      "epoch": 2.5001926782273602,
+      "grad_norm": 0.5856420651001297,
+      "learning_rate": 4.686103135512725e-05,
+      "loss": 0.4845,
+      "step": 811
+    },
+    {
+      "epoch": 2.5032755298651255,
+      "grad_norm": 0.46498562242442115,
+      "learning_rate": 4.677610374884691e-05,
+      "loss": 0.4866,
+      "step": 812
+    },
+    {
+      "epoch": 2.5063583815028903,
+      "grad_norm": 0.2993785576535769,
+      "learning_rate": 4.6691144682142214e-05,
+      "loss": 0.4889,
+      "step": 813
+    },
+    {
+      "epoch": 2.509441233140655,
+      "grad_norm": 0.32779276299010823,
+      "learning_rate": 4.660615454946527e-05,
+      "loss": 0.4887,
+      "step": 814
+    },
+    {
+      "epoch": 2.51252408477842,
+      "grad_norm": 0.48703410433627903,
+      "learning_rate": 4.652113374541239e-05,
+      "loss": 0.4858,
+      "step": 815
+    },
+    {
+      "epoch": 2.515606936416185,
+      "grad_norm": 0.5004031970115542,
+      "learning_rate": 4.64360826647223e-05,
+      "loss": 0.487,
+      "step": 816
+    },
+    {
+      "epoch": 2.51868978805395,
+      "grad_norm": 0.48731183765278296,
+      "learning_rate": 4.6351001702274305e-05,
+      "loss": 0.4833,
+      "step": 817
+    },
+    {
+      "epoch": 2.521772639691715,
+      "grad_norm": 0.3503545758812174,
+      "learning_rate": 4.626589125308643e-05,
+      "loss": 0.4876,
+      "step": 818
+    },
+    {
+      "epoch": 2.52485549132948,
+      "grad_norm": 0.24465846035990946,
+      "learning_rate": 4.618075171231363e-05,
+      "loss": 0.4893,
+      "step": 819
+    },
+    {
+      "epoch": 2.5279383429672446,
+      "grad_norm": 0.30833270786840467,
+      "learning_rate": 4.609558347524589e-05,
+      "loss": 0.4862,
+      "step": 820
+    },
+    {
+      "epoch": 2.5310211946050094,
+      "grad_norm": 0.38607389264508,
+      "learning_rate": 4.601038693730646e-05,
+      "loss": 0.4882,
+      "step": 821
+    },
+    {
+      "epoch": 2.5341040462427746,
+      "grad_norm": 0.4561830183195103,
+      "learning_rate": 4.5925162494049966e-05,
+      "loss": 0.4826,
+      "step": 822
+    },
+    {
+      "epoch": 2.5371868978805394,
+      "grad_norm": 0.42914422422356086,
+      "learning_rate": 4.58399105411606e-05,
+      "loss": 0.4894,
+      "step": 823
+    },
+    {
+      "epoch": 2.5402697495183046,
+      "grad_norm": 0.283710189759877,
+      "learning_rate": 4.575463147445029e-05,
+      "loss": 0.487,
+      "step": 824
+    },
+    {
+      "epoch": 2.5433526011560694,
+      "grad_norm": 0.3092436504519812,
+      "learning_rate": 4.5669325689856824e-05,
+      "loss": 0.4826,
+      "step": 825
+    },
+    {
+      "epoch": 2.546435452793834,
+      "grad_norm": 0.39215676042163716,
+      "learning_rate": 4.558399358344207e-05,
+      "loss": 0.4891,
+      "step": 826
+    },
+    {
+      "epoch": 2.549518304431599,
+      "grad_norm": 0.3754186998804496,
+      "learning_rate": 4.5498635551390064e-05,
+      "loss": 0.486,
+      "step": 827
+    },
+    {
+      "epoch": 2.552601156069364,
+      "grad_norm": 0.3368912609662352,
+      "learning_rate": 4.541325199000525e-05,
+      "loss": 0.4813,
+      "step": 828
+    },
+    {
+      "epoch": 2.555684007707129,
+      "grad_norm": 0.2852375718577191,
+      "learning_rate": 4.532784329571057e-05,
+      "loss": 0.4921,
+      "step": 829
+    },
+    {
+      "epoch": 2.5587668593448942,
+      "grad_norm": 0.41789282160912095,
+      "learning_rate": 4.524240986504569e-05,
+      "loss": 0.4877,
+      "step": 830
+    },
+    {
+      "epoch": 2.561849710982659,
+      "grad_norm": 0.402966863371998,
+      "learning_rate": 4.515695209466507e-05,
+      "loss": 0.4883,
+      "step": 831
+    },
+    {
+      "epoch": 2.564932562620424,
+      "grad_norm": 0.3090730045251245,
+      "learning_rate": 4.507147038133623e-05,
+      "loss": 0.4844,
+      "step": 832
+    },
+    {
+      "epoch": 2.5680154142581886,
+      "grad_norm": 0.3265567472008794,
+      "learning_rate": 4.498596512193782e-05,
+      "loss": 0.4879,
+      "step": 833
+    },
+    {
+      "epoch": 2.571098265895954,
+      "grad_norm": 0.3306111044937201,
+      "learning_rate": 4.490043671345782e-05,
+      "loss": 0.4871,
+      "step": 834
+    },
+    {
+      "epoch": 2.5741811175337186,
+      "grad_norm": 0.33435064420606037,
+      "learning_rate": 4.4814885552991696e-05,
+      "loss": 0.4918,
+      "step": 835
+    },
+    {
+      "epoch": 2.577263969171484,
+      "grad_norm": 0.30449545375884485,
+      "learning_rate": 4.4729312037740534e-05,
+      "loss": 0.4917,
+      "step": 836
+    },
+    {
+      "epoch": 2.5803468208092486,
+      "grad_norm": 0.25794175290735244,
+      "learning_rate": 4.464371656500921e-05,
+      "loss": 0.4911,
+      "step": 837
+    },
+    {
+      "epoch": 2.5834296724470134,
+      "grad_norm": 0.28052212208514027,
+      "learning_rate": 4.455809953220458e-05,
+      "loss": 0.487,
+      "step": 838
+    },
+    {
+      "epoch": 2.586512524084778,
+      "grad_norm": 0.3298572088082483,
+      "learning_rate": 4.4472461336833526e-05,
+      "loss": 0.486,
+      "step": 839
+    },
+    {
+      "epoch": 2.5895953757225434,
+      "grad_norm": 0.39987660673493997,
+      "learning_rate": 4.438680237650127e-05,
+      "loss": 0.4861,
+      "step": 840
+    },
+    {
+      "epoch": 2.592678227360308,
+      "grad_norm": 0.45436998216521973,
+      "learning_rate": 4.430112304890938e-05,
+      "loss": 0.4882,
+      "step": 841
+    },
+    {
+      "epoch": 2.5957610789980734,
+      "grad_norm": 0.4537811504287993,
+      "learning_rate": 4.421542375185403e-05,
+      "loss": 0.4899,
+      "step": 842
+    },
+    {
+      "epoch": 2.598843930635838,
+      "grad_norm": 0.38777980207322177,
+      "learning_rate": 4.412970488322407e-05,
+      "loss": 0.4878,
+      "step": 843
+    },
+    {
+      "epoch": 2.601926782273603,
+      "grad_norm": 0.3493886377264791,
+      "learning_rate": 4.404396684099924e-05,
+      "loss": 0.4886,
+      "step": 844
+    },
+    {
+      "epoch": 2.6050096339113678,
+      "grad_norm": 0.3026755934478254,
+      "learning_rate": 4.395821002324829e-05,
+      "loss": 0.4861,
+      "step": 845
+    },
+    {
+      "epoch": 2.608092485549133,
+      "grad_norm": 0.3862589435209456,
+      "learning_rate": 4.387243482812717e-05,
+      "loss": 0.4884,
+      "step": 846
+    },
+    {
+      "epoch": 2.6111753371868978,
+      "grad_norm": 0.4371892351378419,
+      "learning_rate": 4.378664165387711e-05,
+      "loss": 0.489,
+      "step": 847
+    },
+    {
+      "epoch": 2.614258188824663,
+      "grad_norm": 0.40315213047171555,
+      "learning_rate": 4.3700830898822855e-05,
+      "loss": 0.4883,
+      "step": 848
+    },
+    {
+      "epoch": 2.6173410404624278,
+      "grad_norm": 0.3374872941618157,
+      "learning_rate": 4.361500296137073e-05,
+      "loss": 0.484,
+      "step": 849
+    },
+    {
+      "epoch": 2.6204238921001926,
+      "grad_norm": 0.2938379941142133,
+      "learning_rate": 4.352915824000688e-05,
+      "loss": 0.4843,
+      "step": 850
+    },
+    {
+      "epoch": 2.6235067437379573,
+      "grad_norm": 0.3741406572075112,
+      "learning_rate": 4.344329713329536e-05,
+      "loss": 0.4879,
+      "step": 851
+    },
+    {
+      "epoch": 2.6265895953757226,
+      "grad_norm": 0.31931152338548513,
+      "learning_rate": 4.335742003987627e-05,
+      "loss": 0.4856,
+      "step": 852
+    },
+    {
+      "epoch": 2.6296724470134873,
+      "grad_norm": 0.23409792743931154,
+      "learning_rate": 4.3271527358464e-05,
+      "loss": 0.4856,
+      "step": 853
+    },
+    {
+      "epoch": 2.6327552986512526,
+      "grad_norm": 0.37856007710582756,
+      "learning_rate": 4.318561948784523e-05,
+      "loss": 0.4901,
+      "step": 854
+    },
+    {
+      "epoch": 2.6358381502890174,
+      "grad_norm": 0.4106697824712152,
+      "learning_rate": 4.309969682687724e-05,
+      "loss": 0.4842,
+      "step": 855
+    },
+    {
+      "epoch": 2.638921001926782,
+      "grad_norm": 0.32825464932029724,
+      "learning_rate": 4.301375977448593e-05,
+      "loss": 0.4831,
+      "step": 856
+    },
+    {
+      "epoch": 2.6420038535645474,
+      "grad_norm": 0.2700884530783899,
+      "learning_rate": 4.292780872966401e-05,
+      "loss": 0.4883,
+      "step": 857
+    },
+    {
+      "epoch": 2.645086705202312,
+      "grad_norm": 0.3056096980970238,
+      "learning_rate": 4.284184409146921e-05,
+      "loss": 0.4875,
+      "step": 858
+    },
+    {
+      "epoch": 2.648169556840077,
+      "grad_norm": 0.2816147229319087,
+      "learning_rate": 4.2755866259022305e-05,
+      "loss": 0.4857,
+      "step": 859
+    },
+    {
+      "epoch": 2.651252408477842,
+      "grad_norm": 0.28026747913773264,
+      "learning_rate": 4.2669875631505395e-05,
+      "loss": 0.4859,
+      "step": 860
+    },
+    {
+      "epoch": 2.654335260115607,
+      "grad_norm": 0.2631696686859568,
+      "learning_rate": 4.258387260815992e-05,
+      "loss": 0.487,
+      "step": 861
+    },
+    {
+      "epoch": 2.6574181117533717,
+      "grad_norm": 0.2764223287817884,
+      "learning_rate": 4.2497857588284915e-05,
+      "loss": 0.4886,
+      "step": 862
+    },
+    {
+      "epoch": 2.660500963391137,
+      "grad_norm": 0.2872086269468059,
+      "learning_rate": 4.241183097123509e-05,
+      "loss": 0.4858,
+      "step": 863
+    },
+    {
+      "epoch": 2.6635838150289017,
+      "grad_norm": 0.23287658968443223,
+      "learning_rate": 4.2325793156419035e-05,
+      "loss": 0.4913,
+      "step": 864
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.31053730577851385,
+      "learning_rate": 4.223974454329729e-05,
+      "loss": 0.482,
+      "step": 865
+    },
+    {
+      "epoch": 2.6697495183044317,
+      "grad_norm": 0.24065789576771565,
+      "learning_rate": 4.215368553138056e-05,
+      "loss": 0.4862,
+      "step": 866
+    },
+    {
+      "epoch": 2.6728323699421965,
+      "grad_norm": 0.2689855117734827,
+      "learning_rate": 4.206761652022778e-05,
+      "loss": 0.4844,
+      "step": 867
+    },
+    {
+      "epoch": 2.6759152215799613,
+      "grad_norm": 0.27166925133421277,
+      "learning_rate": 4.198153790944439e-05,
+      "loss": 0.4878,
+      "step": 868
+    },
+    {
+      "epoch": 2.6789980732177265,
+      "grad_norm": 0.2883283999725431,
+      "learning_rate": 4.1895450098680316e-05,
+      "loss": 0.4858,
+      "step": 869
+    },
+    {
+      "epoch": 2.6820809248554913,
+      "grad_norm": 0.2729578583644187,
+      "learning_rate": 4.180935348762826e-05,
+      "loss": 0.4861,
+      "step": 870
+    },
+    {
+      "epoch": 2.685163776493256,
+      "grad_norm": 0.2789107167523073,
+      "learning_rate": 4.1723248476021754e-05,
+      "loss": 0.4865,
+      "step": 871
+    },
+    {
+      "epoch": 2.6882466281310213,
+      "grad_norm": 0.2422587552914448,
+      "learning_rate": 4.163713546363332e-05,
+      "loss": 0.4847,
+      "step": 872
+    },
+    {
+      "epoch": 2.691329479768786,
+      "grad_norm": 0.2478668883959863,
+      "learning_rate": 4.155101485027268e-05,
+      "loss": 0.4879,
+      "step": 873
+    },
+    {
+      "epoch": 2.694412331406551,
+      "grad_norm": 0.224193901504975,
+      "learning_rate": 4.146488703578477e-05,
+      "loss": 0.4854,
+      "step": 874
+    },
+    {
+      "epoch": 2.697495183044316,
+      "grad_norm": 0.22853262909858882,
+      "learning_rate": 4.137875242004802e-05,
+      "loss": 0.4838,
+      "step": 875
+    },
+    {
+      "epoch": 2.700578034682081,
+      "grad_norm": 0.2546449895330431,
+      "learning_rate": 4.129261140297243e-05,
+      "loss": 0.4887,
+      "step": 876
+    },
+    {
+      "epoch": 2.7036608863198457,
+      "grad_norm": 0.27400968228343014,
+      "learning_rate": 4.1206464384497685e-05,
+      "loss": 0.4849,
+      "step": 877
+    },
+    {
+      "epoch": 2.706743737957611,
+      "grad_norm": 0.25547333223360946,
+      "learning_rate": 4.112031176459136e-05,
+      "loss": 0.4813,
+      "step": 878
+    },
+    {
+      "epoch": 2.7098265895953757,
+      "grad_norm": 0.2958590937815924,
+      "learning_rate": 4.103415394324706e-05,
+      "loss": 0.487,
+      "step": 879
+    },
+    {
+      "epoch": 2.7129094412331405,
+      "grad_norm": 0.24685086678561327,
+      "learning_rate": 4.0947991320482455e-05,
+      "loss": 0.4833,
+      "step": 880
+    },
+    {
+      "epoch": 2.7159922928709057,
+      "grad_norm": 0.2603686896229091,
+      "learning_rate": 4.086182429633763e-05,
+      "loss": 0.4857,
+      "step": 881
+    },
+    {
+      "epoch": 2.7190751445086705,
+      "grad_norm": 0.24035103411218198,
+      "learning_rate": 4.077565327087298e-05,
+      "loss": 0.49,
+      "step": 882
+    },
+    {
+      "epoch": 2.7221579961464357,
+      "grad_norm": 0.22931249851506974,
+      "learning_rate": 4.068947864416758e-05,
+      "loss": 0.4858,
+      "step": 883
+    },
+    {
+      "epoch": 2.7252408477842005,
+      "grad_norm": 0.24779519018311094,
+      "learning_rate": 4.0603300816317166e-05,
+      "loss": 0.4867,
+      "step": 884
+    },
+    {
+      "epoch": 2.7283236994219653,
+      "grad_norm": 0.27679625241980743,
+      "learning_rate": 4.051712018743234e-05,
+      "loss": 0.4848,
+      "step": 885
+    },
+    {
+      "epoch": 2.73140655105973,
+      "grad_norm": 0.25177859043952283,
+      "learning_rate": 4.043093715763675e-05,
+      "loss": 0.4852,
+      "step": 886
+    },
+    {
+      "epoch": 2.7344894026974953,
+      "grad_norm": 0.2807694099675048,
+      "learning_rate": 4.034475212706514e-05,
+      "loss": 0.4866,
+      "step": 887
+    },
+    {
+      "epoch": 2.73757225433526,
+      "grad_norm": 0.2733080905396846,
+      "learning_rate": 4.025856549586158e-05,
+      "loss": 0.4876,
+      "step": 888
+    },
+    {
+      "epoch": 2.7406551059730253,
+      "grad_norm": 0.23080162438762253,
+      "learning_rate": 4.017237766417754e-05,
+      "loss": 0.4796,
+      "step": 889
+    },
+    {
+      "epoch": 2.74373795761079,
+      "grad_norm": 0.2504402115994066,
+      "learning_rate": 4.008618903217009e-05,
+      "loss": 0.4834,
+      "step": 890
+    },
+    {
+      "epoch": 2.746820809248555,
+      "grad_norm": 0.20969980637739954,
+      "learning_rate": 4e-05,
+      "loss": 0.487,
+      "step": 891
+    },
+    {
+      "epoch": 2.7499036608863197,
+      "grad_norm": 0.23888791183376637,
+      "learning_rate": 3.9913810967829927e-05,
+      "loss": 0.4878,
+      "step": 892
+    },
+    {
+      "epoch": 2.752986512524085,
+      "grad_norm": 0.23297653178663577,
+      "learning_rate": 3.982762233582248e-05,
+      "loss": 0.4844,
+      "step": 893
+    },
+    {
+      "epoch": 2.7560693641618497,
+      "grad_norm": 0.17047146851845527,
+      "learning_rate": 3.974143450413843e-05,
+      "loss": 0.4839,
+      "step": 894
+    },
+    {
+      "epoch": 2.759152215799615,
+      "grad_norm": 0.2494369213798522,
+      "learning_rate": 3.9655247872934864e-05,
+      "loss": 0.4812,
+      "step": 895
+    },
+    {
+      "epoch": 2.7622350674373797,
+      "grad_norm": 0.23411061792237697,
+      "learning_rate": 3.956906284236326e-05,
+      "loss": 0.4831,
+      "step": 896
+    },
+    {
+      "epoch": 2.7653179190751445,
+      "grad_norm": 0.23847016073602972,
+      "learning_rate": 3.948287981256767e-05,
+      "loss": 0.4863,
+      "step": 897
+    },
+    {
+      "epoch": 2.7684007707129092,
+      "grad_norm": 0.24702286657019246,
+      "learning_rate": 3.9396699183682854e-05,
+      "loss": 0.4864,
+      "step": 898
+    },
+    {
+      "epoch": 2.7714836223506745,
+      "grad_norm": 0.2696090702700198,
+      "learning_rate": 3.931052135583243e-05,
+      "loss": 0.4848,
+      "step": 899
+    },
+    {
+      "epoch": 2.7745664739884393,
+      "grad_norm": 0.21323095257199765,
+      "learning_rate": 3.9224346729127034e-05,
+      "loss": 0.489,
+      "step": 900
+    },
+    {
+      "epoch": 2.7776493256262045,
+      "grad_norm": 0.25816472948458397,
+      "learning_rate": 3.913817570366239e-05,
+      "loss": 0.4804,
+      "step": 901
+    },
+    {
+      "epoch": 2.7807321772639693,
+      "grad_norm": 0.23484155683898814,
+      "learning_rate": 3.905200867951755e-05,
+      "loss": 0.4863,
+      "step": 902
+    },
+    {
+      "epoch": 2.783815028901734,
+      "grad_norm": 0.2286396155476472,
+      "learning_rate": 3.896584605675297e-05,
+      "loss": 0.4843,
+      "step": 903
+    },
+    {
+      "epoch": 2.786897880539499,
+      "grad_norm": 0.2569635869501568,
+      "learning_rate": 3.8879688235408644e-05,
+      "loss": 0.4834,
+      "step": 904
+    },
+    {
+      "epoch": 2.789980732177264,
+      "grad_norm": 0.22618631928155367,
+      "learning_rate": 3.879353561550232e-05,
+      "loss": 0.4845,
+      "step": 905
+    },
+    {
+      "epoch": 2.793063583815029,
+      "grad_norm": 0.2731346233867396,
+      "learning_rate": 3.870738859702758e-05,
+      "loss": 0.4843,
+      "step": 906
+    },
+    {
+      "epoch": 2.796146435452794,
+      "grad_norm": 0.24935340330176858,
+      "learning_rate": 3.8621247579951984e-05,
+      "loss": 0.4837,
+      "step": 907
+    },
+    {
+      "epoch": 2.799229287090559,
+      "grad_norm": 0.26488043800036104,
+      "learning_rate": 3.853511296421524e-05,
+      "loss": 0.4839,
+      "step": 908
+    },
+    {
+      "epoch": 2.8023121387283236,
+      "grad_norm": 0.27905983534505235,
+      "learning_rate": 3.844898514972733e-05,
+      "loss": 0.4873,
+      "step": 909
+    },
+    {
+      "epoch": 2.8053949903660884,
+      "grad_norm": 0.2729417262562729,
+      "learning_rate": 3.836286453636668e-05,
+      "loss": 0.4863,
+      "step": 910
+    },
+    {
+      "epoch": 2.8084778420038536,
+      "grad_norm": 0.21773433316008042,
+      "learning_rate": 3.827675152397826e-05,
+      "loss": 0.4889,
+      "step": 911
+    },
+    {
+      "epoch": 2.8115606936416184,
+      "grad_norm": 0.2634077475665222,
+      "learning_rate": 3.8190646512371754e-05,
+      "loss": 0.4853,
+      "step": 912
+    },
+    {
+      "epoch": 2.8146435452793837,
+      "grad_norm": 0.29655409256589743,
+      "learning_rate": 3.8104549901319704e-05,
+      "loss": 0.4822,
+      "step": 913
+    },
+    {
+      "epoch": 2.8177263969171484,
+      "grad_norm": 0.3026177731332091,
+      "learning_rate": 3.801846209055563e-05,
+      "loss": 0.486,
+      "step": 914
+    },
+    {
+      "epoch": 2.820809248554913,
+      "grad_norm": 0.3179575204031361,
+      "learning_rate": 3.7932383479772224e-05,
+      "loss": 0.484,
+      "step": 915
+    },
+    {
+      "epoch": 2.823892100192678,
+      "grad_norm": 0.2594149197841172,
+      "learning_rate": 3.7846314468619456e-05,
+      "loss": 0.4835,
+      "step": 916
+    },
+    {
+      "epoch": 2.8269749518304432,
+      "grad_norm": 0.2116825583062714,
+      "learning_rate": 3.776025545670272e-05,
+      "loss": 0.4794,
+      "step": 917
+    },
+    {
+      "epoch": 2.830057803468208,
+      "grad_norm": 0.2523898339068698,
+      "learning_rate": 3.767420684358097e-05,
+      "loss": 0.486,
+      "step": 918
+    },
+    {
+      "epoch": 2.8331406551059732,
+      "grad_norm": 0.3365756178714756,
+      "learning_rate": 3.758816902876492e-05,
+      "loss": 0.4846,
+      "step": 919
+    },
+    {
+      "epoch": 2.836223506743738,
+      "grad_norm": 0.28124951649621366,
+      "learning_rate": 3.7502142411715105e-05,
+      "loss": 0.4845,
+      "step": 920
+    },
+    {
+      "epoch": 2.839306358381503,
+      "grad_norm": 0.23801248172986597,
+      "learning_rate": 3.7416127391840094e-05,
+      "loss": 0.4861,
+      "step": 921
+    },
+    {
+      "epoch": 2.8423892100192676,
+      "grad_norm": 0.2525911222477206,
+      "learning_rate": 3.7330124368494625e-05,
+      "loss": 0.4798,
+      "step": 922
+    },
+    {
+      "epoch": 2.845472061657033,
+      "grad_norm": 0.3432867688736019,
+      "learning_rate": 3.72441337409777e-05,
+      "loss": 0.4856,
+      "step": 923
+    },
+    {
+      "epoch": 2.8485549132947976,
+      "grad_norm": 0.2397424795797332,
+      "learning_rate": 3.715815590853081e-05,
+      "loss": 0.4848,
+      "step": 924
+    },
+    {
+      "epoch": 2.851637764932563,
+      "grad_norm": 0.24471548632063347,
+      "learning_rate": 3.707219127033599e-05,
+      "loss": 0.4853,
+      "step": 925
+    },
+    {
+      "epoch": 2.8547206165703276,
+      "grad_norm": 0.2786948339680225,
+      "learning_rate": 3.6986240225514084e-05,
+      "loss": 0.4834,
+      "step": 926
+    },
+    {
+      "epoch": 2.8578034682080924,
+      "grad_norm": 0.21299869857218698,
+      "learning_rate": 3.690030317312277e-05,
+      "loss": 0.4827,
+      "step": 927
+    },
+    {
+      "epoch": 2.860886319845857,
+      "grad_norm": 0.23936681333815235,
+      "learning_rate": 3.6814380512154775e-05,
+      "loss": 0.485,
+      "step": 928
+    },
+    {
+      "epoch": 2.8639691714836224,
+      "grad_norm": 0.34601306428929174,
+      "learning_rate": 3.672847264153602e-05,
+      "loss": 0.4872,
+      "step": 929
+    },
+    {
+      "epoch": 2.867052023121387,
+      "grad_norm": 0.31935875429707156,
+      "learning_rate": 3.664257996012373e-05,
+      "loss": 0.4852,
+      "step": 930
+    },
+    {
+      "epoch": 2.8701348747591524,
+      "grad_norm": 0.25610603842021956,
+      "learning_rate": 3.655670286670465e-05,
+      "loss": 0.4855,
+      "step": 931
+    },
+    {
+      "epoch": 2.873217726396917,
+      "grad_norm": 0.2229137132352307,
+      "learning_rate": 3.647084175999313e-05,
+      "loss": 0.4831,
+      "step": 932
+    },
+    {
+      "epoch": 2.876300578034682,
+      "grad_norm": 0.30626891016065605,
+      "learning_rate": 3.6384997038629286e-05,
+      "loss": 0.4852,
+      "step": 933
+    },
+    {
+      "epoch": 2.8793834296724468,
+      "grad_norm": 0.24851116952780183,
+      "learning_rate": 3.629916910117717e-05,
+      "loss": 0.485,
+      "step": 934
+    },
+    {
+      "epoch": 2.882466281310212,
+      "grad_norm": 0.19819496787466523,
+      "learning_rate": 3.62133583461229e-05,
+      "loss": 0.479,
+      "step": 935
+    },
+    {
+      "epoch": 2.8855491329479768,
+      "grad_norm": 0.22530509202598042,
+      "learning_rate": 3.612756517187284e-05,
+      "loss": 0.4856,
+      "step": 936
+    },
+    {
+      "epoch": 2.888631984585742,
+      "grad_norm": 0.2745286940590315,
+      "learning_rate": 3.604178997675172e-05,
+      "loss": 0.4821,
+      "step": 937
+    },
+    {
+      "epoch": 2.891714836223507,
+      "grad_norm": 0.2483760219614017,
+      "learning_rate": 3.595603315900078e-05,
+      "loss": 0.4855,
+      "step": 938
+    },
+    {
+      "epoch": 2.8947976878612716,
+      "grad_norm": 0.23124747054306113,
+      "learning_rate": 3.587029511677595e-05,
+      "loss": 0.4837,
+      "step": 939
+    },
+    {
+      "epoch": 2.8978805394990363,
+      "grad_norm": 0.27312067066866225,
+      "learning_rate": 3.578457624814598e-05,
+      "loss": 0.4844,
+      "step": 940
+    },
+    {
+      "epoch": 2.9009633911368016,
+      "grad_norm": 0.2511527516553933,
+      "learning_rate": 3.569887695109062e-05,
+      "loss": 0.4806,
+      "step": 941
+    },
+    {
+      "epoch": 2.9040462427745664,
+      "grad_norm": 0.23829581897802005,
+      "learning_rate": 3.561319762349874e-05,
+      "loss": 0.4839,
+      "step": 942
+    },
+    {
+      "epoch": 2.9071290944123316,
+      "grad_norm": 0.2096774001486737,
+      "learning_rate": 3.552753866316649e-05,
+      "loss": 0.4881,
+      "step": 943
+    },
+    {
+      "epoch": 2.9102119460500964,
+      "grad_norm": 0.24340527047999677,
+      "learning_rate": 3.544190046779544e-05,
+      "loss": 0.4825,
+      "step": 944
+    },
+    {
+      "epoch": 2.913294797687861,
+      "grad_norm": 0.2127165492802396,
+      "learning_rate": 3.535628343499079e-05,
+      "loss": 0.483,
+      "step": 945
+    },
+    {
+      "epoch": 2.916377649325626,
+      "grad_norm": 0.20616725196012184,
+      "learning_rate": 3.527068796225947e-05,
+      "loss": 0.4803,
+      "step": 946
+    },
+    {
+      "epoch": 2.919460500963391,
+      "grad_norm": 0.25785811957094146,
+      "learning_rate": 3.518511444700831e-05,
+      "loss": 0.4862,
+      "step": 947
+    },
+    {
+      "epoch": 2.922543352601156,
+      "grad_norm": 0.2770920162651611,
+      "learning_rate": 3.509956328654219e-05,
+      "loss": 0.4855,
+      "step": 948
+    },
+    {
+      "epoch": 2.925626204238921,
+      "grad_norm": 0.28229504977637887,
+      "learning_rate": 3.50140348780622e-05,
+      "loss": 0.4833,
+      "step": 949
+    },
+    {
+      "epoch": 2.928709055876686,
+      "grad_norm": 0.21293331109701055,
+      "learning_rate": 3.4928529618663787e-05,
+      "loss": 0.485,
+      "step": 950
+    },
+    {
+      "epoch": 2.9317919075144507,
+      "grad_norm": 0.2815691721552267,
+      "learning_rate": 3.484304790533494e-05,
+      "loss": 0.4833,
+      "step": 951
+    },
+    {
+      "epoch": 2.934874759152216,
+      "grad_norm": 0.3326621065425569,
+      "learning_rate": 3.4757590134954324e-05,
+      "loss": 0.4872,
+      "step": 952
+    },
+    {
+      "epoch": 2.9379576107899807,
+      "grad_norm": 0.2639000187980299,
+      "learning_rate": 3.467215670428944e-05,
+      "loss": 0.4837,
+      "step": 953
+    },
+    {
+      "epoch": 2.9410404624277455,
+      "grad_norm": 0.3211214007390107,
+      "learning_rate": 3.458674800999477e-05,
+      "loss": 0.4814,
+      "step": 954
+    },
+    {
+      "epoch": 2.9441233140655108,
+      "grad_norm": 0.2861495595090251,
+      "learning_rate": 3.450136444860995e-05,
+      "loss": 0.4846,
+      "step": 955
+    },
+    {
+      "epoch": 2.9472061657032755,
+      "grad_norm": 0.2423985141684349,
+      "learning_rate": 3.441600641655795e-05,
+      "loss": 0.487,
+      "step": 956
+    },
+    {
+      "epoch": 2.9502890173410403,
+      "grad_norm": 0.2441945662089236,
+      "learning_rate": 3.433067431014319e-05,
+      "loss": 0.4819,
+      "step": 957
+    },
+    {
+      "epoch": 2.9533718689788055,
+      "grad_norm": 0.2175740456987027,
+      "learning_rate": 3.4245368525549725e-05,
+      "loss": 0.481,
+      "step": 958
+    },
+    {
+      "epoch": 2.9564547206165703,
+      "grad_norm": 0.24506790207115212,
+      "learning_rate": 3.4160089458839416e-05,
+      "loss": 0.4818,
+      "step": 959
+    },
+    {
+      "epoch": 2.959537572254335,
+      "grad_norm": 0.22533210722978184,
+      "learning_rate": 3.407483750595006e-05,
+      "loss": 0.487,
+      "step": 960
+    },
+    {
+      "epoch": 2.9626204238921003,
+      "grad_norm": 0.21604800541191052,
+      "learning_rate": 3.398961306269355e-05,
+      "loss": 0.4872,
+      "step": 961
+    },
+    {
+      "epoch": 2.965703275529865,
+      "grad_norm": 0.2646427989330409,
+      "learning_rate": 3.390441652475412e-05,
+      "loss": 0.4811,
+      "step": 962
+    },
+    {
+      "epoch": 2.96878612716763,
+      "grad_norm": 0.31179408781271384,
+      "learning_rate": 3.3819248287686386e-05,
+      "loss": 0.4831,
+      "step": 963
+    },
+    {
+      "epoch": 2.971868978805395,
+      "grad_norm": 0.2237743953711824,
+      "learning_rate": 3.3734108746913586e-05,
+      "loss": 0.4869,
+      "step": 964
+    },
+    {
+      "epoch": 2.97495183044316,
+      "grad_norm": 0.2583739805622516,
+      "learning_rate": 3.3648998297725715e-05,
+      "loss": 0.485,
+      "step": 965
+    },
+    {
+      "epoch": 2.9780346820809247,
+      "grad_norm": 0.2946907083307688,
+      "learning_rate": 3.356391733527771e-05,
+      "loss": 0.4815,
+      "step": 966
+    },
+    {
+      "epoch": 2.98111753371869,
+      "grad_norm": 0.20204159431689522,
+      "learning_rate": 3.347886625458762e-05,
+      "loss": 0.4844,
+      "step": 967
+    },
+    {
+      "epoch": 2.9842003853564547,
+      "grad_norm": 0.3207877495094215,
+      "learning_rate": 3.3393845450534746e-05,
+      "loss": 0.4825,
+      "step": 968
+    },
+    {
+      "epoch": 2.9872832369942195,
+      "grad_norm": 0.2585422639938309,
+      "learning_rate": 3.330885531785779e-05,
+      "loss": 0.4842,
+      "step": 969
+    },
+    {
+      "epoch": 2.9903660886319847,
+      "grad_norm": 0.19575655337384346,
+      "learning_rate": 3.322389625115311e-05,
+      "loss": 0.4826,
+      "step": 970
+    },
+    {
+      "epoch": 2.9934489402697495,
+      "grad_norm": 0.24093962541629063,
+      "learning_rate": 3.3138968644872754e-05,
+      "loss": 0.4797,
+      "step": 971
+    },
+    {
+      "epoch": 2.9965317919075143,
+      "grad_norm": 0.2547938326631878,
+      "learning_rate": 3.305407289332279e-05,
+      "loss": 0.483,
+      "step": 972
+    },
+    {
+      "epoch": 2.9996146435452795,
+      "grad_norm": 0.37564929252367313,
+      "learning_rate": 3.296920939066133e-05,
+      "loss": 0.688,
+      "step": 973
+    },
+    {
+      "epoch": 3.0026974951830443,
+      "grad_norm": 0.492986342780779,
+      "learning_rate": 3.2884378530896764e-05,
+      "loss": 0.5603,
+      "step": 974
+    },
+    {
+      "epoch": 3.005780346820809,
+      "grad_norm": 0.45392243644171093,
+      "learning_rate": 3.2799580707885955e-05,
+      "loss": 0.4611,
+      "step": 975
+    },
+    {
+      "epoch": 3.0088631984585743,
+      "grad_norm": 0.354024528080305,
+      "learning_rate": 3.2714816315332346e-05,
+      "loss": 0.4587,
+      "step": 976
+    },
+    {
+      "epoch": 3.011946050096339,
+      "grad_norm": 0.34091993230068657,
+      "learning_rate": 3.26300857467842e-05,
+      "loss": 0.4647,
+      "step": 977
+    },
+    {
+      "epoch": 3.015028901734104,
+      "grad_norm": 0.34379214520363494,
+      "learning_rate": 3.254538939563272e-05,
+      "loss": 0.4613,
+      "step": 978
+    },
+    {
+      "epoch": 3.018111753371869,
+      "grad_norm": 0.3772724277347895,
+      "learning_rate": 3.2460727655110257e-05,
+      "loss": 0.4649,
+      "step": 979
+    },
+    {
+      "epoch": 3.021194605009634,
+      "grad_norm": 0.32986032669870397,
+      "learning_rate": 3.2376100918288445e-05,
+      "loss": 0.4611,
+      "step": 980
+    },
+    {
+      "epoch": 3.0242774566473987,
+      "grad_norm": 0.30049566073185,
+      "learning_rate": 3.229150957807641e-05,
+      "loss": 0.4576,
+      "step": 981
+    },
+    {
+      "epoch": 3.027360308285164,
+      "grad_norm": 0.3377290082159356,
+      "learning_rate": 3.2206954027218976e-05,
+      "loss": 0.457,
+      "step": 982
+    },
+    {
+      "epoch": 3.0304431599229287,
+      "grad_norm": 0.2862196076955671,
+      "learning_rate": 3.212243465829476e-05,
+      "loss": 0.457,
+      "step": 983
+    },
+    {
+      "epoch": 3.0335260115606935,
+      "grad_norm": 0.26037506735076504,
+      "learning_rate": 3.203795186371439e-05,
+      "loss": 0.457,
+      "step": 984
+    },
+    {
+      "epoch": 3.0366088631984587,
+      "grad_norm": 0.2439637483643575,
+      "learning_rate": 3.19535060357187e-05,
+      "loss": 0.4587,
+      "step": 985
+    },
+    {
+      "epoch": 3.0396917148362235,
+      "grad_norm": 0.27553395620700866,
+      "learning_rate": 3.186909756637689e-05,
+      "loss": 0.4574,
+      "step": 986
+    },
+    {
+      "epoch": 3.0427745664739883,
+      "grad_norm": 0.2628380698092703,
+      "learning_rate": 3.1784726847584723e-05,
+      "loss": 0.4554,
+      "step": 987
+    },
+    {
+      "epoch": 3.0458574181117535,
+      "grad_norm": 0.23428542473842123,
+      "learning_rate": 3.170039427106267e-05,
+      "loss": 0.4582,
+      "step": 988
+    },
+    {
+      "epoch": 3.0489402697495183,
+      "grad_norm": 0.21610133122013445,
+      "learning_rate": 3.161610022835414e-05,
+      "loss": 0.4561,
+      "step": 989
+    },
+    {
+      "epoch": 3.052023121387283,
+      "grad_norm": 0.22610133332019758,
+      "learning_rate": 3.153184511082359e-05,
+      "loss": 0.4571,
+      "step": 990
+    },
+    {
+      "epoch": 3.0551059730250483,
+      "grad_norm": 0.21951922401267213,
+      "learning_rate": 3.144762930965478e-05,
+      "loss": 0.458,
+      "step": 991
+    },
+    {
+      "epoch": 3.058188824662813,
+      "grad_norm": 0.23200973107687622,
+      "learning_rate": 3.1363453215848956e-05,
+      "loss": 0.4603,
+      "step": 992
+    },
+    {
+      "epoch": 3.061271676300578,
+      "grad_norm": 0.22086231438517664,
+      "learning_rate": 3.127931722022297e-05,
+      "loss": 0.4572,
+      "step": 993
+    },
+    {
+      "epoch": 3.064354527938343,
+      "grad_norm": 0.21696808813685436,
+      "learning_rate": 3.119522171340752e-05,
+      "loss": 0.4559,
+      "step": 994
+    },
+    {
+      "epoch": 3.067437379576108,
+      "grad_norm": 0.1959764538032747,
+      "learning_rate": 3.111116708584533e-05,
+      "loss": 0.4602,
+      "step": 995
+    },
+    {
+      "epoch": 3.0705202312138726,
+      "grad_norm": 0.24036888250548819,
+      "learning_rate": 3.102715372778931e-05,
+      "loss": 0.4592,
+      "step": 996
+    },
+    {
+      "epoch": 3.073603082851638,
+      "grad_norm": 0.18825952562251913,
+      "learning_rate": 3.094318202930074e-05,
+      "loss": 0.4561,
+      "step": 997
+    },
+    {
+      "epoch": 3.0766859344894026,
+      "grad_norm": 0.21740285418614763,
+      "learning_rate": 3.085925238024755e-05,
+      "loss": 0.459,
+      "step": 998
+    },
+    {
+      "epoch": 3.0797687861271674,
+      "grad_norm": 0.21104658453208688,
+      "learning_rate": 3.07753651703024e-05,
+      "loss": 0.4579,
+      "step": 999
+    },
+    {
+      "epoch": 3.0828516377649327,
+      "grad_norm": 0.2345007689307205,
+      "learning_rate": 3.069152078894089e-05,
+      "loss": 0.459,
+      "step": 1000
+    },
+    {
+      "epoch": 3.0859344894026974,
+      "grad_norm": 0.21488598311905108,
+      "learning_rate": 3.06077196254398e-05,
+      "loss": 0.4554,
+      "step": 1001
+    },
+    {
+      "epoch": 3.089017341040462,
+      "grad_norm": 0.1911838973949256,
+      "learning_rate": 3.0523962068875234e-05,
+      "loss": 0.4609,
+      "step": 1002
+    },
+    {
+      "epoch": 3.0921001926782274,
+      "grad_norm": 0.20798725961673994,
+      "learning_rate": 3.0440248508120898e-05,
+      "loss": 0.4559,
+      "step": 1003
+    },
+    {
+      "epoch": 3.0951830443159922,
+      "grad_norm": 0.18879223309267432,
+      "learning_rate": 3.0356579331846156e-05,
+      "loss": 0.457,
+      "step": 1004
+    },
+    {
+      "epoch": 3.098265895953757,
+      "grad_norm": 0.17962271217627118,
+      "learning_rate": 3.0272954928514343e-05,
+      "loss": 0.455,
+      "step": 1005
+    },
+    {
+      "epoch": 3.1013487475915222,
+      "grad_norm": 0.17853040172762857,
+      "learning_rate": 3.0189375686380905e-05,
+      "loss": 0.4584,
+      "step": 1006
+    },
+    {
+      "epoch": 3.104431599229287,
+      "grad_norm": 0.17893685747473798,
+      "learning_rate": 3.0105841993491593e-05,
+      "loss": 0.4587,
+      "step": 1007
+    },
+    {
+      "epoch": 3.1075144508670522,
+      "grad_norm": 0.19514452850726854,
+      "learning_rate": 3.0022354237680752e-05,
+      "loss": 0.4547,
+      "step": 1008
+    },
+    {
+      "epoch": 3.110597302504817,
+      "grad_norm": 0.19379317401955565,
+      "learning_rate": 2.9938912806569372e-05,
+      "loss": 0.4549,
+      "step": 1009
+    },
+    {
+      "epoch": 3.113680154142582,
+      "grad_norm": 0.19147508322542164,
+      "learning_rate": 2.9855518087563388e-05,
+      "loss": 0.4614,
+      "step": 1010
+    },
+    {
+      "epoch": 3.1167630057803466,
+      "grad_norm": 0.2193617098874827,
+      "learning_rate": 2.9772170467851863e-05,
+      "loss": 0.4565,
+      "step": 1011
+    },
+    {
+      "epoch": 3.119845857418112,
+      "grad_norm": 0.19653400142114508,
+      "learning_rate": 2.9688870334405183e-05,
+      "loss": 0.4583,
+      "step": 1012
+    },
+    {
+      "epoch": 3.1229287090558766,
+      "grad_norm": 0.16533246810792449,
+      "learning_rate": 2.9605618073973274e-05,
+      "loss": 0.4602,
+      "step": 1013
+    },
+    {
+      "epoch": 3.126011560693642,
+      "grad_norm": 0.20805762221547505,
+      "learning_rate": 2.9522414073083767e-05,
+      "loss": 0.4582,
+      "step": 1014
+    },
+    {
+      "epoch": 3.1290944123314066,
+      "grad_norm": 0.20019248131520406,
+      "learning_rate": 2.943925871804026e-05,
+      "loss": 0.4591,
+      "step": 1015
+    },
+    {
+      "epoch": 3.1321772639691714,
+      "grad_norm": 0.17961760976150198,
+      "learning_rate": 2.9356152394920486e-05,
+      "loss": 0.4554,
+      "step": 1016
+    },
+    {
+      "epoch": 3.135260115606936,
+      "grad_norm": 0.20008558840376753,
+      "learning_rate": 2.9273095489574502e-05,
+      "loss": 0.4587,
+      "step": 1017
+    },
+    {
+      "epoch": 3.1383429672447014,
+      "grad_norm": 0.21707760329573628,
+      "learning_rate": 2.919008838762299e-05,
+      "loss": 0.4601,
+      "step": 1018
+    },
+    {
+      "epoch": 3.141425818882466,
+      "grad_norm": 0.16876877551083475,
+      "learning_rate": 2.9107131474455337e-05,
+      "loss": 0.4581,
+      "step": 1019
+    },
+    {
+      "epoch": 3.1445086705202314,
+      "grad_norm": 0.20097923140349075,
+      "learning_rate": 2.9024225135227947e-05,
+      "loss": 0.4575,
+      "step": 1020
+    },
+    {
+      "epoch": 3.147591522157996,
+      "grad_norm": 0.19821433741556183,
+      "learning_rate": 2.894136975486241e-05,
+      "loss": 0.4577,
+      "step": 1021
+    },
+    {
+      "epoch": 3.150674373795761,
+      "grad_norm": 0.16106505072198857,
+      "learning_rate": 2.885856571804369e-05,
+      "loss": 0.4542,
+      "step": 1022
+    },
+    {
+      "epoch": 3.153757225433526,
+      "grad_norm": 0.22519408145714895,
+      "learning_rate": 2.8775813409218448e-05,
+      "loss": 0.4585,
+      "step": 1023
+    },
+    {
+      "epoch": 3.156840077071291,
+      "grad_norm": 0.17524335974077918,
+      "learning_rate": 2.8693113212593107e-05,
+      "loss": 0.4597,
+      "step": 1024
+    },
+    {
+      "epoch": 3.159922928709056,
+      "grad_norm": 0.21951856756934532,
+      "learning_rate": 2.861046551213217e-05,
+      "loss": 0.4588,
+      "step": 1025
+    },
+    {
+      "epoch": 3.163005780346821,
+      "grad_norm": 0.24055212248845456,
+      "learning_rate": 2.8527870691556404e-05,
+      "loss": 0.456,
+      "step": 1026
+    },
+    {
+      "epoch": 3.166088631984586,
+      "grad_norm": 0.1977198687379575,
+      "learning_rate": 2.8445329134341055e-05,
+      "loss": 0.458,
+      "step": 1027
+    },
+    {
+      "epoch": 3.1691714836223506,
+      "grad_norm": 0.21686815916971341,
+      "learning_rate": 2.8362841223714074e-05,
+      "loss": 0.4581,
+      "step": 1028
+    },
+    {
+      "epoch": 3.172254335260116,
+      "grad_norm": 0.20768343546982562,
+      "learning_rate": 2.8280407342654375e-05,
+      "loss": 0.4579,
+      "step": 1029
+    },
+    {
+      "epoch": 3.1753371868978806,
+      "grad_norm": 0.18244846062893144,
+      "learning_rate": 2.8198027873889988e-05,
+      "loss": 0.4544,
+      "step": 1030
+    },
+    {
+      "epoch": 3.1784200385356454,
+      "grad_norm": 0.19316173395905548,
+      "learning_rate": 2.8115703199896303e-05,
+      "loss": 0.4587,
+      "step": 1031
+    },
+    {
+      "epoch": 3.1815028901734106,
+      "grad_norm": 0.16077441743746157,
+      "learning_rate": 2.8033433702894345e-05,
+      "loss": 0.4571,
+      "step": 1032
+    },
+    {
+      "epoch": 3.1845857418111754,
+      "grad_norm": 0.17803077305079368,
+      "learning_rate": 2.795121976484892e-05,
+      "loss": 0.4583,
+      "step": 1033
+    },
+    {
+      "epoch": 3.18766859344894,
+      "grad_norm": 0.16662238794723377,
+      "learning_rate": 2.7869061767466915e-05,
+      "loss": 0.4576,
+      "step": 1034
+    },
+    {
+      "epoch": 3.1907514450867054,
+      "grad_norm": 0.16200538468140283,
+      "learning_rate": 2.778696009219548e-05,
+      "loss": 0.4557,
+      "step": 1035
+    },
+    {
+      "epoch": 3.19383429672447,
+      "grad_norm": 0.18300760261714777,
+      "learning_rate": 2.7704915120220264e-05,
+      "loss": 0.4622,
+      "step": 1036
+    },
+    {
+      "epoch": 3.196917148362235,
+      "grad_norm": 0.15356757169963633,
+      "learning_rate": 2.7622927232463666e-05,
+      "loss": 0.4604,
+      "step": 1037
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.15928517995572616,
+      "learning_rate": 2.7540996809583017e-05,
+      "loss": 0.4529,
+      "step": 1038
+    },
+    {
+      "epoch": 3.203082851637765,
+      "grad_norm": 0.16551528152652456,
+      "learning_rate": 2.745912423196891e-05,
+      "loss": 0.4591,
+      "step": 1039
+    },
+    {
+      "epoch": 3.2061657032755297,
+      "grad_norm": 0.17215703482174827,
+      "learning_rate": 2.7377309879743304e-05,
+      "loss": 0.4588,
+      "step": 1040
+    },
+    {
+      "epoch": 3.209248554913295,
+      "grad_norm": 0.1518332218041993,
+      "learning_rate": 2.7295554132757877e-05,
+      "loss": 0.4537,
+      "step": 1041
+    },
+    {
+      "epoch": 3.2123314065510598,
+      "grad_norm": 0.17440332608301698,
+      "learning_rate": 2.721385737059218e-05,
+      "loss": 0.4601,
+      "step": 1042
+    },
+    {
+      "epoch": 3.2154142581888245,
+      "grad_norm": 0.14561328541264035,
+      "learning_rate": 2.7132219972551906e-05,
+      "loss": 0.4597,
+      "step": 1043
+    },
+    {
+      "epoch": 3.2184971098265898,
+      "grad_norm": 0.1731249196327213,
+      "learning_rate": 2.7050642317667164e-05,
+      "loss": 0.4597,
+      "step": 1044
+    },
+    {
+      "epoch": 3.2215799614643545,
+      "grad_norm": 0.17075835079331075,
+      "learning_rate": 2.6969124784690653e-05,
+      "loss": 0.4605,
+      "step": 1045
+    },
+    {
+      "epoch": 3.2246628131021193,
+      "grad_norm": 0.17205757538175617,
+      "learning_rate": 2.688766775209594e-05,
+      "loss": 0.4617,
+      "step": 1046
+    },
+    {
+      "epoch": 3.2277456647398846,
+      "grad_norm": 0.1845633347672156,
+      "learning_rate": 2.6806271598075695e-05,
+      "loss": 0.4568,
+      "step": 1047
+    },
+    {
+      "epoch": 3.2308285163776493,
+      "grad_norm": 0.17511048185029482,
+      "learning_rate": 2.6724936700539955e-05,
+      "loss": 0.4582,
+      "step": 1048
+    },
+    {
+      "epoch": 3.233911368015414,
+      "grad_norm": 0.17495094883621043,
+      "learning_rate": 2.6643663437114343e-05,
+      "loss": 0.4555,
+      "step": 1049
+    },
+    {
+      "epoch": 3.2369942196531793,
+      "grad_norm": 0.1904716916680108,
+      "learning_rate": 2.6562452185138323e-05,
+      "loss": 0.4591,
+      "step": 1050
+    },
+    {
+      "epoch": 3.240077071290944,
+      "grad_norm": 0.18299600026968044,
+      "learning_rate": 2.6481303321663437e-05,
+      "loss": 0.4563,
+      "step": 1051
+    },
+    {
+      "epoch": 3.243159922928709,
+      "grad_norm": 0.17711381479050994,
+      "learning_rate": 2.640021722345159e-05,
+      "loss": 0.4596,
+      "step": 1052
+    },
+    {
+      "epoch": 3.246242774566474,
+      "grad_norm": 0.17847378414554854,
+      "learning_rate": 2.6319194266973256e-05,
+      "loss": 0.4557,
+      "step": 1053
+    },
+    {
+      "epoch": 3.249325626204239,
+      "grad_norm": 0.1879329920173585,
+      "learning_rate": 2.6238234828405763e-05,
+      "loss": 0.4554,
+      "step": 1054
+    },
+    {
+      "epoch": 3.2524084778420037,
+      "grad_norm": 0.19733442293120765,
+      "learning_rate": 2.6157339283631555e-05,
+      "loss": 0.4603,
+      "step": 1055
+    },
+    {
+      "epoch": 3.255491329479769,
+      "grad_norm": 0.1695302327977651,
+      "learning_rate": 2.6076508008236384e-05,
+      "loss": 0.4593,
+      "step": 1056
+    },
+    {
+      "epoch": 3.2585741811175337,
+      "grad_norm": 0.18946228814563748,
+      "learning_rate": 2.5995741377507634e-05,
+      "loss": 0.4576,
+      "step": 1057
+    },
+    {
+      "epoch": 3.2616570327552985,
+      "grad_norm": 0.1798769160926996,
+      "learning_rate": 2.591503976643254e-05,
+      "loss": 0.4561,
+      "step": 1058
+    },
+    {
+      "epoch": 3.2647398843930637,
+      "grad_norm": 0.1870937324805083,
+      "learning_rate": 2.5834403549696498e-05,
+      "loss": 0.4592,
+      "step": 1059
+    },
+    {
+      "epoch": 3.2678227360308285,
+      "grad_norm": 0.16596057939719594,
+      "learning_rate": 2.5753833101681238e-05,
+      "loss": 0.4617,
+      "step": 1060
+    },
+    {
+      "epoch": 3.2709055876685933,
+      "grad_norm": 0.18252237129906898,
+      "learning_rate": 2.5673328796463173e-05,
+      "loss": 0.4577,
+      "step": 1061
+    },
+    {
+      "epoch": 3.2739884393063585,
+      "grad_norm": 0.1885964677349154,
+      "learning_rate": 2.5592891007811594e-05,
+      "loss": 0.4601,
+      "step": 1062
+    },
+    {
+      "epoch": 3.2770712909441233,
+      "grad_norm": 0.15843970935017654,
+      "learning_rate": 2.5512520109186977e-05,
+      "loss": 0.4559,
+      "step": 1063
+    },
+    {
+      "epoch": 3.280154142581888,
+      "grad_norm": 0.19413962956113548,
+      "learning_rate": 2.543221647373927e-05,
+      "loss": 0.4565,
+      "step": 1064
+    },
+    {
+      "epoch": 3.2832369942196533,
+      "grad_norm": 0.19342148517856397,
+      "learning_rate": 2.5351980474306083e-05,
+      "loss": 0.4629,
+      "step": 1065
+    },
+    {
+      "epoch": 3.286319845857418,
+      "grad_norm": 0.18328450288676756,
+      "learning_rate": 2.527181248341103e-05,
+      "loss": 0.4608,
+      "step": 1066
+    },
+    {
+      "epoch": 3.289402697495183,
+      "grad_norm": 0.20653762657184196,
+      "learning_rate": 2.5191712873261943e-05,
+      "loss": 0.4605,
+      "step": 1067
+    },
+    {
+      "epoch": 3.292485549132948,
+      "grad_norm": 0.17924258207322835,
+      "learning_rate": 2.5111682015749212e-05,
+      "loss": 0.4606,
+      "step": 1068
+    },
+    {
+      "epoch": 3.295568400770713,
+      "grad_norm": 0.16735081398506932,
+      "learning_rate": 2.5031720282443977e-05,
+      "loss": 0.4579,
+      "step": 1069
+    },
+    {
+      "epoch": 3.2986512524084777,
+      "grad_norm": 0.16868615279351504,
+      "learning_rate": 2.495182804459648e-05,
+      "loss": 0.4552,
+      "step": 1070
+    },
+    {
+      "epoch": 3.301734104046243,
+      "grad_norm": 0.14827968066722017,
+      "learning_rate": 2.4872005673134307e-05,
+      "loss": 0.4623,
+      "step": 1071
+    },
+    {
+      "epoch": 3.3048169556840077,
+      "grad_norm": 0.1696540496843515,
+      "learning_rate": 2.4792253538660626e-05,
+      "loss": 0.4569,
+      "step": 1072
+    },
+    {
+      "epoch": 3.3078998073217725,
+      "grad_norm": 0.15959995545162392,
+      "learning_rate": 2.4712572011452537e-05,
+      "loss": 0.4558,
+      "step": 1073
+    },
+    {
+      "epoch": 3.3109826589595377,
+      "grad_norm": 0.17712479977617046,
+      "learning_rate": 2.4632961461459294e-05,
+      "loss": 0.458,
+      "step": 1074
+    },
+    {
+      "epoch": 3.3140655105973025,
+      "grad_norm": 0.1592651483430736,
+      "learning_rate": 2.4553422258300676e-05,
+      "loss": 0.4572,
+      "step": 1075
+    },
+    {
+      "epoch": 3.3171483622350673,
+      "grad_norm": 0.18617319958941894,
+      "learning_rate": 2.447395477126514e-05,
+      "loss": 0.4594,
+      "step": 1076
+    },
+    {
+      "epoch": 3.3202312138728325,
+      "grad_norm": 0.1607459295122554,
+      "learning_rate": 2.4394559369308208e-05,
+      "loss": 0.4577,
+      "step": 1077
+    },
+    {
+      "epoch": 3.3233140655105973,
+      "grad_norm": 0.16626179011426007,
+      "learning_rate": 2.431523642105072e-05,
+      "loss": 0.4592,
+      "step": 1078
+    },
+    {
+      "epoch": 3.326396917148362,
+      "grad_norm": 0.18437478413906744,
+      "learning_rate": 2.4235986294777103e-05,
+      "loss": 0.4564,
+      "step": 1079
+    },
+    {
+      "epoch": 3.3294797687861273,
+      "grad_norm": 0.15743988312200322,
+      "learning_rate": 2.4156809358433728e-05,
+      "loss": 0.4521,
+      "step": 1080
+    },
+    {
+      "epoch": 3.332562620423892,
+      "grad_norm": 0.16858300831854253,
+      "learning_rate": 2.4077705979627128e-05,
+      "loss": 0.4613,
+      "step": 1081
+    },
+    {
+      "epoch": 3.335645472061657,
+      "grad_norm": 0.14754811108095806,
+      "learning_rate": 2.3998676525622296e-05,
+      "loss": 0.4619,
+      "step": 1082
+    },
+    {
+      "epoch": 3.338728323699422,
+      "grad_norm": 0.17587166719177644,
+      "learning_rate": 2.391972136334106e-05,
+      "loss": 0.4579,
+      "step": 1083
+    },
+    {
+      "epoch": 3.341811175337187,
+      "grad_norm": 0.15532045300351083,
+      "learning_rate": 2.3840840859360248e-05,
+      "loss": 0.4601,
+      "step": 1084
+    },
+    {
+      "epoch": 3.3448940269749516,
+      "grad_norm": 0.1721679692549858,
+      "learning_rate": 2.376203537991015e-05,
+      "loss": 0.457,
+      "step": 1085
+    },
+    {
+      "epoch": 3.347976878612717,
+      "grad_norm": 0.14918518833041025,
+      "learning_rate": 2.368330529087266e-05,
+      "loss": 0.4635,
+      "step": 1086
+    },
+    {
+      "epoch": 3.3510597302504816,
+      "grad_norm": 0.1677810139901804,
+      "learning_rate": 2.3604650957779667e-05,
+      "loss": 0.4569,
+      "step": 1087
+    },
+    {
+      "epoch": 3.3541425818882464,
+      "grad_norm": 0.16113106788670006,
+      "learning_rate": 2.352607274581135e-05,
+      "loss": 0.4575,
+      "step": 1088
+    },
+    {
+      "epoch": 3.3572254335260117,
+      "grad_norm": 0.1740201560460167,
+      "learning_rate": 2.3447571019794438e-05,
+      "loss": 0.4596,
+      "step": 1089
+    },
+    {
+      "epoch": 3.3603082851637764,
+      "grad_norm": 0.1496381511985011,
+      "learning_rate": 2.3369146144200575e-05,
+      "loss": 0.4585,
+      "step": 1090
+    },
+    {
+      "epoch": 3.3633911368015417,
+      "grad_norm": 0.13954555829711576,
+      "learning_rate": 2.3290798483144607e-05,
+      "loss": 0.4617,
+      "step": 1091
+    },
+    {
+      "epoch": 3.3664739884393065,
+      "grad_norm": 0.14287358240609022,
+      "learning_rate": 2.321252840038284e-05,
+      "loss": 0.4594,
+      "step": 1092
+    },
+    {
+      "epoch": 3.3695568400770712,
+      "grad_norm": 0.1385208025366805,
+      "learning_rate": 2.3134336259311457e-05,
+      "loss": 0.4596,
+      "step": 1093
+    },
+    {
+      "epoch": 3.372639691714836,
+      "grad_norm": 0.14844841891368812,
+      "learning_rate": 2.3056222422964684e-05,
+      "loss": 0.4563,
+      "step": 1094
+    },
+    {
+      "epoch": 3.3757225433526012,
+      "grad_norm": 0.13222566344488543,
+      "learning_rate": 2.297818725401331e-05,
+      "loss": 0.4568,
+      "step": 1095
+    },
+    {
+      "epoch": 3.378805394990366,
+      "grad_norm": 0.16129510137362083,
+      "learning_rate": 2.290023111476277e-05,
+      "loss": 0.457,
+      "step": 1096
+    },
+    {
+      "epoch": 3.3818882466281313,
+      "grad_norm": 0.14958155580573967,
+      "learning_rate": 2.282235436715163e-05,
+      "loss": 0.4559,
+      "step": 1097
+    },
+    {
+      "epoch": 3.384971098265896,
+      "grad_norm": 0.17115251040081172,
+      "learning_rate": 2.274455737274987e-05,
+      "loss": 0.4603,
+      "step": 1098
+    },
+    {
+      "epoch": 3.388053949903661,
+      "grad_norm": 0.17346563183701436,
+      "learning_rate": 2.2666840492757132e-05,
+      "loss": 0.4621,
+      "step": 1099
+    },
+    {
+      "epoch": 3.3911368015414256,
+      "grad_norm": 0.14543612736039485,
+      "learning_rate": 2.2589204088001138e-05,
+      "loss": 0.4597,
+      "step": 1100
+    },
+    {
+      "epoch": 3.394219653179191,
+      "grad_norm": 0.16806073877299738,
+      "learning_rate": 2.251164851893597e-05,
+      "loss": 0.4611,
+      "step": 1101
+    },
+    {
+      "epoch": 3.3973025048169556,
+      "grad_norm": 0.14709734818037748,
+      "learning_rate": 2.243417414564042e-05,
+      "loss": 0.4599,
+      "step": 1102
+    },
+    {
+      "epoch": 3.400385356454721,
+      "grad_norm": 0.1655198448166463,
+      "learning_rate": 2.235678132781624e-05,
+      "loss": 0.4585,
+      "step": 1103
+    },
+    {
+      "epoch": 3.4034682080924856,
+      "grad_norm": 0.18448674069153864,
+      "learning_rate": 2.2279470424786607e-05,
+      "loss": 0.4564,
+      "step": 1104
+    },
+    {
+      "epoch": 3.4065510597302504,
+      "grad_norm": 0.17500442360948873,
+      "learning_rate": 2.22022417954943e-05,
+      "loss": 0.4585,
+      "step": 1105
+    },
+    {
+      "epoch": 3.409633911368015,
+      "grad_norm": 0.16951807800261065,
+      "learning_rate": 2.212509579850022e-05,
+      "loss": 0.4609,
+      "step": 1106
+    },
+    {
+      "epoch": 3.4127167630057804,
+      "grad_norm": 0.14795882562393395,
+      "learning_rate": 2.2048032791981515e-05,
+      "loss": 0.4591,
+      "step": 1107
+    },
+    {
+      "epoch": 3.415799614643545,
+      "grad_norm": 0.18615497340332252,
+      "learning_rate": 2.1971053133730092e-05,
+      "loss": 0.4596,
+      "step": 1108
+    },
+    {
+      "epoch": 3.4188824662813104,
+      "grad_norm": 0.1473346637940834,
+      "learning_rate": 2.1894157181150817e-05,
+      "loss": 0.4578,
+      "step": 1109
+    },
+    {
+      "epoch": 3.421965317919075,
+      "grad_norm": 0.168652212536844,
+      "learning_rate": 2.181734529125998e-05,
+      "loss": 0.4601,
+      "step": 1110
+    },
+    {
+      "epoch": 3.42504816955684,
+      "grad_norm": 0.17042637031813804,
+      "learning_rate": 2.1740617820683568e-05,
+      "loss": 0.4577,
+      "step": 1111
+    },
+    {
+      "epoch": 3.4281310211946048,
+      "grad_norm": 0.14122415541376404,
+      "learning_rate": 2.1663975125655633e-05,
+      "loss": 0.455,
+      "step": 1112
+    },
+    {
+      "epoch": 3.43121387283237,
+      "grad_norm": 0.17153317629862405,
+      "learning_rate": 2.1587417562016578e-05,
+      "loss": 0.4582,
+      "step": 1113
+    },
+    {
+      "epoch": 3.434296724470135,
+      "grad_norm": 0.14369139411336637,
+      "learning_rate": 2.15109454852116e-05,
+      "loss": 0.4578,
+      "step": 1114
+    },
+    {
+      "epoch": 3.4373795761079,
+      "grad_norm": 0.1792915914534855,
+      "learning_rate": 2.1434559250288977e-05,
+      "loss": 0.4589,
+      "step": 1115
+    },
+    {
+      "epoch": 3.440462427745665,
+      "grad_norm": 0.1675767742836398,
+      "learning_rate": 2.135825921189846e-05,
+      "loss": 0.455,
+      "step": 1116
+    },
+    {
+      "epoch": 3.4435452793834296,
+      "grad_norm": 0.15534809361121651,
+      "learning_rate": 2.1282045724289545e-05,
+      "loss": 0.4559,
+      "step": 1117
+    },
+    {
+      "epoch": 3.4466281310211944,
+      "grad_norm": 0.1780806380906026,
+      "learning_rate": 2.1205919141309942e-05,
+      "loss": 0.4579,
+      "step": 1118
+    },
+    {
+      "epoch": 3.4497109826589596,
+      "grad_norm": 0.1786658809222873,
+      "learning_rate": 2.112987981640387e-05,
+      "loss": 0.4554,
+      "step": 1119
+    },
+    {
+      "epoch": 3.4527938342967244,
+      "grad_norm": 0.1939709216807228,
+      "learning_rate": 2.1053928102610362e-05,
+      "loss": 0.4577,
+      "step": 1120
+    },
+    {
+      "epoch": 3.4558766859344896,
+      "grad_norm": 0.18277723697994191,
+      "learning_rate": 2.09780643525618e-05,
+      "loss": 0.4532,
+      "step": 1121
+    },
+    {
+      "epoch": 3.4589595375722544,
+      "grad_norm": 0.1955838220437729,
+      "learning_rate": 2.0902288918482042e-05,
+      "loss": 0.458,
+      "step": 1122
+    },
+    {
+      "epoch": 3.462042389210019,
+      "grad_norm": 0.15160261240878362,
+      "learning_rate": 2.082660215218499e-05,
+      "loss": 0.4567,
+      "step": 1123
+    },
+    {
+      "epoch": 3.4651252408477844,
+      "grad_norm": 0.20758760499216244,
+      "learning_rate": 2.075100440507283e-05,
+      "loss": 0.4593,
+      "step": 1124
+    },
+    {
+      "epoch": 3.468208092485549,
+      "grad_norm": 0.1925710428856299,
+      "learning_rate": 2.067549602813446e-05,
+      "loss": 0.4552,
+      "step": 1125
+    },
+    {
+      "epoch": 3.471290944123314,
+      "grad_norm": 0.18994350863383008,
+      "learning_rate": 2.0600077371943863e-05,
+      "loss": 0.4563,
+      "step": 1126
+    },
+    {
+      "epoch": 3.474373795761079,
+      "grad_norm": 0.17861422749901273,
+      "learning_rate": 2.0524748786658463e-05,
+      "loss": 0.4587,
+      "step": 1127
+    },
+    {
+      "epoch": 3.477456647398844,
+      "grad_norm": 0.2077927838797709,
+      "learning_rate": 2.0449510622017437e-05,
+      "loss": 0.4633,
+      "step": 1128
+    },
+    {
+      "epoch": 3.4805394990366088,
+      "grad_norm": 0.1930225083019618,
+      "learning_rate": 2.0374363227340243e-05,
+      "loss": 0.4587,
+      "step": 1129
+    },
+    {
+      "epoch": 3.483622350674374,
+      "grad_norm": 0.21352046680187967,
+      "learning_rate": 2.029930695152481e-05,
+      "loss": 0.4583,
+      "step": 1130
+    },
+    {
+      "epoch": 3.4867052023121388,
+      "grad_norm": 0.1721633233392436,
+      "learning_rate": 2.0224342143046128e-05,
+      "loss": 0.4602,
+      "step": 1131
+    },
+    {
+      "epoch": 3.4897880539499035,
+      "grad_norm": 0.22022549034341085,
+      "learning_rate": 2.014946914995443e-05,
+      "loss": 0.4555,
+      "step": 1132
+    },
+    {
+      "epoch": 3.4928709055876688,
+      "grad_norm": 0.14675007693278738,
+      "learning_rate": 2.007468831987369e-05,
+      "loss": 0.4583,
+      "step": 1133
+    },
+    {
+      "epoch": 3.4959537572254336,
+      "grad_norm": 0.24535572020452398,
+      "learning_rate": 2.0000000000000012e-05,
+      "loss": 0.4591,
+      "step": 1134
+    },
+    {
+      "epoch": 3.4990366088631983,
+      "grad_norm": 0.15054113031961827,
+      "learning_rate": 1.9925404537099924e-05,
+      "loss": 0.4549,
+      "step": 1135
+    },
+    {
+      "epoch": 3.502119460500963,
+      "grad_norm": 0.19694646121967324,
+      "learning_rate": 1.9850902277508894e-05,
+      "loss": 0.4569,
+      "step": 1136
+    },
+    {
+      "epoch": 3.5052023121387283,
+      "grad_norm": 0.17274731615114983,
+      "learning_rate": 1.9776493567129643e-05,
+      "loss": 0.456,
+      "step": 1137
+    },
+    {
+      "epoch": 3.508285163776493,
+      "grad_norm": 0.21807349421858177,
+      "learning_rate": 1.9702178751430563e-05,
+      "loss": 0.4599,
+      "step": 1138
+    },
+    {
+      "epoch": 3.5113680154142584,
+      "grad_norm": 0.1632405258067428,
+      "learning_rate": 1.9627958175444065e-05,
+      "loss": 0.4601,
+      "step": 1139
+    },
+    {
+      "epoch": 3.514450867052023,
+      "grad_norm": 0.21577735803695933,
+      "learning_rate": 1.9553832183765075e-05,
+      "loss": 0.4599,
+      "step": 1140
+    },
+    {
+      "epoch": 3.517533718689788,
+      "grad_norm": 0.16170580231653672,
+      "learning_rate": 1.9479801120549352e-05,
+      "loss": 0.4607,
+      "step": 1141
+    },
+    {
+      "epoch": 3.520616570327553,
+      "grad_norm": 0.1887197950432579,
+      "learning_rate": 1.9405865329511938e-05,
+      "loss": 0.4588,
+      "step": 1142
+    },
+    {
+      "epoch": 3.523699421965318,
+      "grad_norm": 0.18825373682176638,
+      "learning_rate": 1.9332025153925486e-05,
+      "loss": 0.4601,
+      "step": 1143
+    },
+    {
+      "epoch": 3.5267822736030827,
+      "grad_norm": 0.19775661743543468,
+      "learning_rate": 1.9258280936618794e-05,
+      "loss": 0.4579,
+      "step": 1144
+    },
+    {
+      "epoch": 3.529865125240848,
+      "grad_norm": 0.18866279625440852,
+      "learning_rate": 1.9184633019975073e-05,
+      "loss": 0.4597,
+      "step": 1145
+    },
+    {
+      "epoch": 3.5329479768786127,
+      "grad_norm": 0.16562255719242155,
+      "learning_rate": 1.911108174593047e-05,
+      "loss": 0.4565,
+      "step": 1146
+    },
+    {
+      "epoch": 3.5360308285163775,
+      "grad_norm": 0.1963366126086293,
+      "learning_rate": 1.9037627455972414e-05,
+      "loss": 0.4596,
+      "step": 1147
+    },
+    {
+      "epoch": 3.5391136801541427,
+      "grad_norm": 0.14816688602740055,
+      "learning_rate": 1.8964270491138087e-05,
+      "loss": 0.4555,
+      "step": 1148
+    },
+    {
+      "epoch": 3.5421965317919075,
+      "grad_norm": 0.2203670911303755,
+      "learning_rate": 1.8891011192012736e-05,
+      "loss": 0.4554,
+      "step": 1149
+    },
+    {
+      "epoch": 3.5452793834296723,
+      "grad_norm": 0.1588450174156007,
+      "learning_rate": 1.881784989872822e-05,
+      "loss": 0.4576,
+      "step": 1150
+    },
+    {
+      "epoch": 3.5483622350674375,
+      "grad_norm": 0.2011726633180639,
+      "learning_rate": 1.8744786950961354e-05,
+      "loss": 0.4535,
+      "step": 1151
+    },
+    {
+      "epoch": 3.5514450867052023,
+      "grad_norm": 0.18064889678392101,
+      "learning_rate": 1.867182268793236e-05,
+      "loss": 0.4585,
+      "step": 1152
+    },
+    {
+      "epoch": 3.554527938342967,
+      "grad_norm": 0.17674551997689075,
+      "learning_rate": 1.8598957448403242e-05,
+      "loss": 0.4564,
+      "step": 1153
+    },
+    {
+      "epoch": 3.5576107899807323,
+      "grad_norm": 0.1616685690835829,
+      "learning_rate": 1.8526191570676296e-05,
+      "loss": 0.4569,
+      "step": 1154
+    },
+    {
+      "epoch": 3.560693641618497,
+      "grad_norm": 0.16535937370751339,
+      "learning_rate": 1.84535253925925e-05,
+      "loss": 0.4571,
+      "step": 1155
+    },
+    {
+      "epoch": 3.563776493256262,
+      "grad_norm": 0.1913085824758526,
+      "learning_rate": 1.8380959251529873e-05,
+      "loss": 0.4551,
+      "step": 1156
+    },
+    {
+      "epoch": 3.566859344894027,
+      "grad_norm": 0.14735456166837058,
+      "learning_rate": 1.8308493484402086e-05,
+      "loss": 0.4612,
+      "step": 1157
+    },
+    {
+      "epoch": 3.569942196531792,
+      "grad_norm": 0.21592848997525524,
+      "learning_rate": 1.8236128427656692e-05,
+      "loss": 0.4567,
+      "step": 1158
+    },
+    {
+      "epoch": 3.5730250481695567,
+      "grad_norm": 0.14056237803572807,
+      "learning_rate": 1.8163864417273728e-05,
+      "loss": 0.4579,
+      "step": 1159
+    },
+    {
+      "epoch": 3.576107899807322,
+      "grad_norm": 0.20921685795095712,
+      "learning_rate": 1.8091701788764024e-05,
+      "loss": 0.4543,
+      "step": 1160
+    },
+    {
+      "epoch": 3.5791907514450867,
+      "grad_norm": 0.16504254062556628,
+      "learning_rate": 1.8019640877167763e-05,
+      "loss": 0.4578,
+      "step": 1161
+    },
+    {
+      "epoch": 3.5822736030828515,
+      "grad_norm": 0.19376241076998563,
+      "learning_rate": 1.794768201705285e-05,
+      "loss": 0.4602,
+      "step": 1162
+    },
+    {
+      "epoch": 3.5853564547206167,
+      "grad_norm": 0.1504409364891568,
+      "learning_rate": 1.7875825542513402e-05,
+      "loss": 0.4582,
+      "step": 1163
+    },
+    {
+      "epoch": 3.5884393063583815,
+      "grad_norm": 0.21873536061261592,
+      "learning_rate": 1.780407178716813e-05,
+      "loss": 0.4567,
+      "step": 1164
+    },
+    {
+      "epoch": 3.5915221579961463,
+      "grad_norm": 0.16375771064645817,
+      "learning_rate": 1.7732421084158887e-05,
+      "loss": 0.4568,
+      "step": 1165
+    },
+    {
+      "epoch": 3.5946050096339115,
+      "grad_norm": 0.15558645891106893,
+      "learning_rate": 1.7660873766149e-05,
+      "loss": 0.4597,
+      "step": 1166
+    },
+    {
+      "epoch": 3.5976878612716763,
+      "grad_norm": 0.15176277514598843,
+      "learning_rate": 1.7589430165321902e-05,
+      "loss": 0.4549,
+      "step": 1167
+    },
+    {
+      "epoch": 3.6007707129094415,
+      "grad_norm": 0.18759612208890505,
+      "learning_rate": 1.751809061337938e-05,
+      "loss": 0.4584,
+      "step": 1168
+    },
+    {
+      "epoch": 3.6038535645472063,
+      "grad_norm": 0.14328133563445636,
+      "learning_rate": 1.7446855441540204e-05,
+      "loss": 0.4558,
+      "step": 1169
+    },
+    {
+      "epoch": 3.606936416184971,
+      "grad_norm": 0.18277805110234197,
+      "learning_rate": 1.7375724980538465e-05,
+      "loss": 0.4607,
+      "step": 1170
+    },
+    {
+      "epoch": 3.610019267822736,
+      "grad_norm": 0.1643119330159651,
+      "learning_rate": 1.7304699560622147e-05,
+      "loss": 0.4568,
+      "step": 1171
+    },
+    {
+      "epoch": 3.613102119460501,
+      "grad_norm": 0.1620078268156403,
+      "learning_rate": 1.7233779511551532e-05,
+      "loss": 0.4564,
+      "step": 1172
+    },
+    {
+      "epoch": 3.616184971098266,
+      "grad_norm": 0.19002722863231578,
+      "learning_rate": 1.716296516259767e-05,
+      "loss": 0.4569,
+      "step": 1173
+    },
+    {
+      "epoch": 3.619267822736031,
+      "grad_norm": 0.1976694486190197,
+      "learning_rate": 1.7092256842540883e-05,
+      "loss": 0.4583,
+      "step": 1174
+    },
+    {
+      "epoch": 3.622350674373796,
+      "grad_norm": 0.15963785862104055,
+      "learning_rate": 1.7021654879669175e-05,
+      "loss": 0.4619,
+      "step": 1175
+    },
+    {
+      "epoch": 3.6254335260115607,
+      "grad_norm": 0.18414298673715002,
+      "learning_rate": 1.695115960177679e-05,
+      "loss": 0.456,
+      "step": 1176
+    },
+    {
+      "epoch": 3.6285163776493254,
+      "grad_norm": 0.14356178702290384,
+      "learning_rate": 1.6880771336162633e-05,
+      "loss": 0.4548,
+      "step": 1177
+    },
+    {
+      "epoch": 3.6315992292870907,
+      "grad_norm": 0.18777819506219637,
+      "learning_rate": 1.681049040962879e-05,
+      "loss": 0.4553,
+      "step": 1178
+    },
+    {
+      "epoch": 3.6346820809248555,
+      "grad_norm": 0.14751720341519717,
+      "learning_rate": 1.6740317148478932e-05,
+      "loss": 0.4601,
+      "step": 1179
+    },
+    {
+      "epoch": 3.6377649325626207,
+      "grad_norm": 0.17441323991112753,
+      "learning_rate": 1.6670251878516933e-05,
+      "loss": 0.4619,
+      "step": 1180
+    },
+    {
+      "epoch": 3.6408477842003855,
+      "grad_norm": 0.15225392071649851,
+      "learning_rate": 1.660029492504519e-05,
+      "loss": 0.457,
+      "step": 1181
+    },
+    {
+      "epoch": 3.6439306358381502,
+      "grad_norm": 0.15019930564119924,
+      "learning_rate": 1.653044661286329e-05,
+      "loss": 0.4585,
+      "step": 1182
+    },
+    {
+      "epoch": 3.647013487475915,
+      "grad_norm": 0.14635047782627306,
+      "learning_rate": 1.646070726626636e-05,
+      "loss": 0.4576,
+      "step": 1183
+    },
+    {
+      "epoch": 3.6500963391136803,
+      "grad_norm": 0.14237903282704836,
+      "learning_rate": 1.639107720904366e-05,
+      "loss": 0.4581,
+      "step": 1184
+    },
+    {
+      "epoch": 3.653179190751445,
+      "grad_norm": 0.14414783151789698,
+      "learning_rate": 1.6321556764476965e-05,
+      "loss": 0.4582,
+      "step": 1185
+    },
+    {
+      "epoch": 3.6562620423892103,
+      "grad_norm": 0.15140245554562762,
+      "learning_rate": 1.6252146255339223e-05,
+      "loss": 0.457,
+      "step": 1186
+    },
+    {
+      "epoch": 3.659344894026975,
+      "grad_norm": 0.1367220452530055,
+      "learning_rate": 1.6182846003892878e-05,
+      "loss": 0.4592,
+      "step": 1187
+    },
+    {
+      "epoch": 3.66242774566474,
+      "grad_norm": 0.17027730997917362,
+      "learning_rate": 1.6113656331888563e-05,
+      "loss": 0.4577,
+      "step": 1188
+    },
+    {
+      "epoch": 3.6655105973025046,
+      "grad_norm": 0.14385488573712285,
+      "learning_rate": 1.6044577560563412e-05,
+      "loss": 0.454,
+      "step": 1189
+    },
+    {
+      "epoch": 3.66859344894027,
+      "grad_norm": 0.16602936435980745,
+      "learning_rate": 1.5975610010639734e-05,
+      "loss": 0.4586,
+      "step": 1190
+    },
+    {
+      "epoch": 3.6716763005780346,
+      "grad_norm": 0.17744112177748064,
+      "learning_rate": 1.5906754002323427e-05,
+      "loss": 0.4539,
+      "step": 1191
+    },
+    {
+      "epoch": 3.6747591522158,
+      "grad_norm": 0.1656846102175362,
+      "learning_rate": 1.583800985530249e-05,
+      "loss": 0.4574,
+      "step": 1192
+    },
+    {
+      "epoch": 3.6778420038535646,
+      "grad_norm": 0.1653877662054782,
+      "learning_rate": 1.5769377888745626e-05,
+      "loss": 0.4547,
+      "step": 1193
+    },
+    {
+      "epoch": 3.6809248554913294,
+      "grad_norm": 0.1620948602930088,
+      "learning_rate": 1.5700858421300655e-05,
+      "loss": 0.4585,
+      "step": 1194
+    },
+    {
+      "epoch": 3.684007707129094,
+      "grad_norm": 0.16845365823481778,
+      "learning_rate": 1.5632451771093112e-05,
+      "loss": 0.4572,
+      "step": 1195
+    },
+    {
+      "epoch": 3.6870905587668594,
+      "grad_norm": 0.17500212921427186,
+      "learning_rate": 1.55641582557247e-05,
+      "loss": 0.4595,
+      "step": 1196
+    },
+    {
+      "epoch": 3.690173410404624,
+      "grad_norm": 0.1467619010985275,
+      "learning_rate": 1.5495978192271887e-05,
+      "loss": 0.4605,
+      "step": 1197
+    },
+    {
+      "epoch": 3.6932562620423894,
+      "grad_norm": 0.1602762637144391,
+      "learning_rate": 1.5427911897284392e-05,
+      "loss": 0.4583,
+      "step": 1198
+    },
+    {
+      "epoch": 3.696339113680154,
+      "grad_norm": 0.16073967708951695,
+      "learning_rate": 1.5359959686783736e-05,
+      "loss": 0.4556,
+      "step": 1199
+    },
+    {
+      "epoch": 3.699421965317919,
+      "grad_norm": 0.14314500313570916,
+      "learning_rate": 1.5292121876261723e-05,
+      "loss": 0.4558,
+      "step": 1200
+    },
+    {
+      "epoch": 3.702504816955684,
+      "grad_norm": 0.1543476713488105,
+      "learning_rate": 1.5224398780679069e-05,
+      "loss": 0.4605,
+      "step": 1201
+    },
+    {
+      "epoch": 3.705587668593449,
+      "grad_norm": 0.14394740791587998,
+      "learning_rate": 1.5156790714463814e-05,
+      "loss": 0.457,
+      "step": 1202
+    },
+    {
+      "epoch": 3.708670520231214,
+      "grad_norm": 0.1529865687786838,
+      "learning_rate": 1.5089297991510044e-05,
+      "loss": 0.4604,
+      "step": 1203
+    },
+    {
+      "epoch": 3.711753371868979,
+      "grad_norm": 0.1418143541562142,
+      "learning_rate": 1.5021920925176208e-05,
+      "loss": 0.4566,
+      "step": 1204
+    },
+    {
+      "epoch": 3.714836223506744,
+      "grad_norm": 0.1518341674432377,
+      "learning_rate": 1.4954659828283857e-05,
+      "loss": 0.4603,
+      "step": 1205
+    },
+    {
+      "epoch": 3.7179190751445086,
+      "grad_norm": 0.1399279372322184,
+      "learning_rate": 1.4887515013116067e-05,
+      "loss": 0.4564,
+      "step": 1206
+    },
+    {
+      "epoch": 3.7210019267822734,
+      "grad_norm": 0.16139309054948844,
+      "learning_rate": 1.482048679141606e-05,
+      "loss": 0.4568,
+      "step": 1207
+    },
+    {
+      "epoch": 3.7240847784200386,
+      "grad_norm": 0.1359764373724039,
+      "learning_rate": 1.475357547438573e-05,
+      "loss": 0.4592,
+      "step": 1208
+    },
+    {
+      "epoch": 3.7271676300578034,
+      "grad_norm": 0.15809021016631364,
+      "learning_rate": 1.468678137268421e-05,
+      "loss": 0.4592,
+      "step": 1209
+    },
+    {
+      "epoch": 3.7302504816955686,
+      "grad_norm": 0.14410285341264376,
+      "learning_rate": 1.462010479642638e-05,
+      "loss": 0.458,
+      "step": 1210
+    },
+    {
+      "epoch": 3.7333333333333334,
+      "grad_norm": 0.12524520056787203,
+      "learning_rate": 1.4553546055181511e-05,
+      "loss": 0.461,
+      "step": 1211
+    },
+    {
+      "epoch": 3.736416184971098,
+      "grad_norm": 0.14918734835790373,
+      "learning_rate": 1.4487105457971757e-05,
+      "loss": 0.4552,
+      "step": 1212
+    },
+    {
+      "epoch": 3.739499036608863,
+      "grad_norm": 0.11819270807476978,
+      "learning_rate": 1.442078331327077e-05,
+      "loss": 0.4535,
+      "step": 1213
+    },
+    {
+      "epoch": 3.742581888246628,
+      "grad_norm": 0.1343788231190205,
+      "learning_rate": 1.4354579929002235e-05,
+      "loss": 0.458,
+      "step": 1214
+    },
+    {
+      "epoch": 3.745664739884393,
+      "grad_norm": 0.11997737439806964,
+      "learning_rate": 1.4288495612538427e-05,
+      "loss": 0.4575,
+      "step": 1215
+    },
+    {
+      "epoch": 3.748747591522158,
+      "grad_norm": 0.1392960815572695,
+      "learning_rate": 1.4222530670698853e-05,
+      "loss": 0.4601,
+      "step": 1216
+    },
+    {
+      "epoch": 3.751830443159923,
+      "grad_norm": 0.13299411789981813,
+      "learning_rate": 1.4156685409748714e-05,
+      "loss": 0.4585,
+      "step": 1217
+    },
+    {
+      "epoch": 3.7549132947976878,
+      "grad_norm": 0.13155466563407478,
+      "learning_rate": 1.4090960135397644e-05,
+      "loss": 0.4561,
+      "step": 1218
+    },
+    {
+      "epoch": 3.7579961464354525,
+      "grad_norm": 0.1337890257966772,
+      "learning_rate": 1.4025355152798113e-05,
+      "loss": 0.4568,
+      "step": 1219
+    },
+    {
+      "epoch": 3.7610789980732178,
+      "grad_norm": 0.13449434330164853,
+      "learning_rate": 1.3959870766544153e-05,
+      "loss": 0.4576,
+      "step": 1220
+    },
+    {
+      "epoch": 3.7641618497109826,
+      "grad_norm": 0.13797568623011458,
+      "learning_rate": 1.3894507280669828e-05,
+      "loss": 0.4554,
+      "step": 1221
+    },
+    {
+      "epoch": 3.767244701348748,
+      "grad_norm": 0.13246989415082205,
+      "learning_rate": 1.3829264998647945e-05,
+      "loss": 0.4593,
+      "step": 1222
+    },
+    {
+      "epoch": 3.7703275529865126,
+      "grad_norm": 0.14147501767267615,
+      "learning_rate": 1.3764144223388502e-05,
+      "loss": 0.4597,
+      "step": 1223
+    },
+    {
+      "epoch": 3.7734104046242773,
+      "grad_norm": 0.1651172541299993,
+      "learning_rate": 1.369914525723746e-05,
+      "loss": 0.4565,
+      "step": 1224
+    },
+    {
+      "epoch": 3.776493256262042,
+      "grad_norm": 0.13954523176810946,
+      "learning_rate": 1.3634268401975148e-05,
+      "loss": 0.4571,
+      "step": 1225
+    },
+    {
+      "epoch": 3.7795761078998074,
+      "grad_norm": 0.14666029620150922,
+      "learning_rate": 1.3569513958814993e-05,
+      "loss": 0.4606,
+      "step": 1226
+    },
+    {
+      "epoch": 3.782658959537572,
+      "grad_norm": 0.15475181834914362,
+      "learning_rate": 1.3504882228402095e-05,
+      "loss": 0.4533,
+      "step": 1227
+    },
+    {
+      "epoch": 3.7857418111753374,
+      "grad_norm": 0.15438381147917543,
+      "learning_rate": 1.3440373510811768e-05,
+      "loss": 0.4584,
+      "step": 1228
+    },
+    {
+      "epoch": 3.788824662813102,
+      "grad_norm": 0.15009770882139262,
+      "learning_rate": 1.3375988105548246e-05,
+      "loss": 0.4569,
+      "step": 1229
+    },
+    {
+      "epoch": 3.791907514450867,
+      "grad_norm": 0.1352193569562934,
+      "learning_rate": 1.3311726311543213e-05,
+      "loss": 0.4561,
+      "step": 1230
+    },
+    {
+      "epoch": 3.7949903660886317,
+      "grad_norm": 0.13474476558402831,
+      "learning_rate": 1.3247588427154474e-05,
+      "loss": 0.4597,
+      "step": 1231
+    },
+    {
+      "epoch": 3.798073217726397,
+      "grad_norm": 0.14463592082671103,
+      "learning_rate": 1.3183574750164497e-05,
+      "loss": 0.4605,
+      "step": 1232
+    },
+    {
+      "epoch": 3.8011560693641617,
+      "grad_norm": 0.12965561795963806,
+      "learning_rate": 1.3119685577779105e-05,
+      "loss": 0.4603,
+      "step": 1233
+    },
+    {
+      "epoch": 3.804238921001927,
+      "grad_norm": 0.15680400192347,
+      "learning_rate": 1.3055921206626061e-05,
+      "loss": 0.4559,
+      "step": 1234
+    },
+    {
+      "epoch": 3.8073217726396917,
+      "grad_norm": 0.12573093154131584,
+      "learning_rate": 1.2992281932753704e-05,
+      "loss": 0.4556,
+      "step": 1235
+    },
+    {
+      "epoch": 3.8104046242774565,
+      "grad_norm": 0.13185548413226023,
+      "learning_rate": 1.2928768051629526e-05,
+      "loss": 0.4595,
+      "step": 1236
+    },
+    {
+      "epoch": 3.8134874759152213,
+      "grad_norm": 0.12757665912613708,
+      "learning_rate": 1.2865379858138889e-05,
+      "loss": 0.4549,
+      "step": 1237
+    },
+    {
+      "epoch": 3.8165703275529865,
+      "grad_norm": 0.14199014689347056,
+      "learning_rate": 1.280211764658355e-05,
+      "loss": 0.458,
+      "step": 1238
+    },
+    {
+      "epoch": 3.8196531791907513,
+      "grad_norm": 0.11187146428971001,
+      "learning_rate": 1.2738981710680439e-05,
+      "loss": 0.4567,
+      "step": 1239
+    },
+    {
+      "epoch": 3.8227360308285165,
+      "grad_norm": 0.1357605015866328,
+      "learning_rate": 1.2675972343560106e-05,
+      "loss": 0.4576,
+      "step": 1240
+    },
+    {
+      "epoch": 3.8258188824662813,
+      "grad_norm": 0.1338822783361523,
+      "learning_rate": 1.2613089837765534e-05,
+      "loss": 0.4557,
+      "step": 1241
+    },
+    {
+      "epoch": 3.828901734104046,
+      "grad_norm": 0.11751005972066009,
+      "learning_rate": 1.2550334485250661e-05,
+      "loss": 0.456,
+      "step": 1242
+    },
+    {
+      "epoch": 3.8319845857418113,
+      "grad_norm": 0.12335966930445133,
+      "learning_rate": 1.2487706577379108e-05,
+      "loss": 0.4565,
+      "step": 1243
+    },
+    {
+      "epoch": 3.835067437379576,
+      "grad_norm": 0.12883332516061122,
+      "learning_rate": 1.242520640492277e-05,
+      "loss": 0.4577,
+      "step": 1244
+    },
+    {
+      "epoch": 3.838150289017341,
+      "grad_norm": 0.12773557268712435,
+      "learning_rate": 1.23628342580605e-05,
+      "loss": 0.4603,
+      "step": 1245
+    },
+    {
+      "epoch": 3.841233140655106,
+      "grad_norm": 0.12137962775715713,
+      "learning_rate": 1.2300590426376716e-05,
+      "loss": 0.4594,
+      "step": 1246
+    },
+    {
+      "epoch": 3.844315992292871,
+      "grad_norm": 0.12873102261007582,
+      "learning_rate": 1.2238475198860122e-05,
+      "loss": 0.4545,
+      "step": 1247
+    },
+    {
+      "epoch": 3.8473988439306357,
+      "grad_norm": 0.11902380257709168,
+      "learning_rate": 1.217648886390232e-05,
+      "loss": 0.4552,
+      "step": 1248
+    },
+    {
+      "epoch": 3.850481695568401,
+      "grad_norm": 0.11139839162591711,
+      "learning_rate": 1.2114631709296485e-05,
+      "loss": 0.4589,
+      "step": 1249
+    },
+    {
+      "epoch": 3.8535645472061657,
+      "grad_norm": 0.14167370735273932,
+      "learning_rate": 1.2052904022236037e-05,
+      "loss": 0.4553,
+      "step": 1250
+    },
+    {
+      "epoch": 3.8566473988439305,
+      "grad_norm": 0.12606450539286848,
+      "learning_rate": 1.1991306089313261e-05,
+      "loss": 0.4591,
+      "step": 1251
+    },
+    {
+      "epoch": 3.8597302504816957,
+      "grad_norm": 0.11224700468530312,
+      "learning_rate": 1.1929838196518064e-05,
+      "loss": 0.458,
+      "step": 1252
+    },
+    {
+      "epoch": 3.8628131021194605,
+      "grad_norm": 0.13708233924811727,
+      "learning_rate": 1.1868500629236533e-05,
+      "loss": 0.4586,
+      "step": 1253
+    },
+    {
+      "epoch": 3.8658959537572253,
+      "grad_norm": 0.13023528433635065,
+      "learning_rate": 1.1807293672249757e-05,
+      "loss": 0.459,
+      "step": 1254
+    },
+    {
+      "epoch": 3.8689788053949905,
+      "grad_norm": 0.12631430770585264,
+      "learning_rate": 1.1746217609732348e-05,
+      "loss": 0.4601,
+      "step": 1255
+    },
+    {
+      "epoch": 3.8720616570327553,
+      "grad_norm": 0.1304708396049557,
+      "learning_rate": 1.1685272725251244e-05,
+      "loss": 0.4545,
+      "step": 1256
+    },
+    {
+      "epoch": 3.87514450867052,
+      "grad_norm": 0.1469259602181826,
+      "learning_rate": 1.1624459301764288e-05,
+      "loss": 0.4585,
+      "step": 1257
+    },
+    {
+      "epoch": 3.8782273603082853,
+      "grad_norm": 0.12324779303466817,
+      "learning_rate": 1.1563777621619038e-05,
+      "loss": 0.4583,
+      "step": 1258
+    },
+    {
+      "epoch": 3.88131021194605,
+      "grad_norm": 0.12214316725359158,
+      "learning_rate": 1.1503227966551313e-05,
+      "loss": 0.4586,
+      "step": 1259
+    },
+    {
+      "epoch": 3.884393063583815,
+      "grad_norm": 0.13714791051468003,
+      "learning_rate": 1.1442810617684046e-05,
+      "loss": 0.4578,
+      "step": 1260
+    },
+    {
+      "epoch": 3.88747591522158,
+      "grad_norm": 0.11653974873062173,
+      "learning_rate": 1.1382525855525808e-05,
+      "loss": 0.4555,
+      "step": 1261
+    },
+    {
+      "epoch": 3.890558766859345,
+      "grad_norm": 0.11446517128817475,
+      "learning_rate": 1.1322373959969654e-05,
+      "loss": 0.4569,
+      "step": 1262
+    },
+    {
+      "epoch": 3.89364161849711,
+      "grad_norm": 0.12594588661402423,
+      "learning_rate": 1.1262355210291709e-05,
+      "loss": 0.4577,
+      "step": 1263
+    },
+    {
+      "epoch": 3.896724470134875,
+      "grad_norm": 0.1115661293597501,
+      "learning_rate": 1.120246988514996e-05,
+      "loss": 0.4557,
+      "step": 1264
+    },
+    {
+      "epoch": 3.8998073217726397,
+      "grad_norm": 0.11393598207330236,
+      "learning_rate": 1.1142718262582904e-05,
+      "loss": 0.4576,
+      "step": 1265
+    },
+    {
+      "epoch": 3.9028901734104045,
+      "grad_norm": 0.12334631289097314,
+      "learning_rate": 1.1083100620008284e-05,
+      "loss": 0.4601,
+      "step": 1266
+    },
+    {
+      "epoch": 3.9059730250481697,
+      "grad_norm": 0.11504477029583844,
+      "learning_rate": 1.102361723422181e-05,
+      "loss": 0.4577,
+      "step": 1267
+    },
+    {
+      "epoch": 3.9090558766859345,
+      "grad_norm": 0.12300189351723968,
+      "learning_rate": 1.0964268381395816e-05,
+      "loss": 0.4585,
+      "step": 1268
+    },
+    {
+      "epoch": 3.9121387283236997,
+      "grad_norm": 0.12329112191325854,
+      "learning_rate": 1.0905054337078051e-05,
+      "loss": 0.4567,
+      "step": 1269
+    },
+    {
+      "epoch": 3.9152215799614645,
+      "grad_norm": 0.12032824881333183,
+      "learning_rate": 1.084597537619037e-05,
+      "loss": 0.4587,
+      "step": 1270
+    },
+    {
+      "epoch": 3.9183044315992293,
+      "grad_norm": 0.11877650745881703,
+      "learning_rate": 1.0787031773027458e-05,
+      "loss": 0.4563,
+      "step": 1271
+    },
+    {
+      "epoch": 3.921387283236994,
+      "grad_norm": 0.12368252509022372,
+      "learning_rate": 1.0728223801255519e-05,
+      "loss": 0.4582,
+      "step": 1272
+    },
+    {
+      "epoch": 3.9244701348747593,
+      "grad_norm": 0.12155121671993799,
+      "learning_rate": 1.0669551733911092e-05,
+      "loss": 0.4576,
+      "step": 1273
+    },
+    {
+      "epoch": 3.927552986512524,
+      "grad_norm": 0.12342259934353209,
+      "learning_rate": 1.0611015843399674e-05,
+      "loss": 0.4567,
+      "step": 1274
+    },
+    {
+      "epoch": 3.9306358381502893,
+      "grad_norm": 0.11257333694926747,
+      "learning_rate": 1.0552616401494594e-05,
+      "loss": 0.4562,
+      "step": 1275
+    },
+    {
+      "epoch": 3.933718689788054,
+      "grad_norm": 0.12509555521993196,
+      "learning_rate": 1.0494353679335588e-05,
+      "loss": 0.4574,
+      "step": 1276
+    },
+    {
+      "epoch": 3.936801541425819,
+      "grad_norm": 0.1289221446144822,
+      "learning_rate": 1.0436227947427677e-05,
+      "loss": 0.4598,
+      "step": 1277
+    },
+    {
+      "epoch": 3.9398843930635836,
+      "grad_norm": 0.11508249318016571,
+      "learning_rate": 1.0378239475639823e-05,
+      "loss": 0.4543,
+      "step": 1278
+    },
+    {
+      "epoch": 3.942967244701349,
+      "grad_norm": 0.12644156744693982,
+      "learning_rate": 1.0320388533203736e-05,
+      "loss": 0.4589,
+      "step": 1279
+    },
+    {
+      "epoch": 3.9460500963391136,
+      "grad_norm": 0.12726637344053127,
+      "learning_rate": 1.0262675388712582e-05,
+      "loss": 0.4547,
+      "step": 1280
+    },
+    {
+      "epoch": 3.949132947976879,
+      "grad_norm": 0.12584327601272044,
+      "learning_rate": 1.0205100310119773e-05,
+      "loss": 0.4603,
+      "step": 1281
+    },
+    {
+      "epoch": 3.9522157996146436,
+      "grad_norm": 0.1211357673645595,
+      "learning_rate": 1.0147663564737669e-05,
+      "loss": 0.4544,
+      "step": 1282
+    },
+    {
+      "epoch": 3.9552986512524084,
+      "grad_norm": 0.1178628681978078,
+      "learning_rate": 1.0090365419236389e-05,
+      "loss": 0.4537,
+      "step": 1283
+    },
+    {
+      "epoch": 3.958381502890173,
+      "grad_norm": 0.1220433244110075,
+      "learning_rate": 1.003320613964256e-05,
+      "loss": 0.4581,
+      "step": 1284
+    },
+    {
+      "epoch": 3.9614643545279384,
+      "grad_norm": 0.11529554191758529,
+      "learning_rate": 9.976185991338073e-06,
+      "loss": 0.4548,
+      "step": 1285
+    },
+    {
+      "epoch": 3.964547206165703,
+      "grad_norm": 0.12576783226419594,
+      "learning_rate": 9.919305239058832e-06,
+      "loss": 0.4571,
+      "step": 1286
+    },
+    {
+      "epoch": 3.9676300578034684,
+      "grad_norm": 0.12616907173539701,
+      "learning_rate": 9.862564146893571e-06,
+      "loss": 0.4592,
+      "step": 1287
+    },
+    {
+      "epoch": 3.9707129094412332,
+      "grad_norm": 0.13973351824367156,
+      "learning_rate": 9.805962978282607e-06,
+      "loss": 0.4608,
+      "step": 1288
+    },
+    {
+      "epoch": 3.973795761078998,
+      "grad_norm": 0.1359938475002537,
+      "learning_rate": 9.749501996016564e-06,
+      "loss": 0.4622,
+      "step": 1289
+    },
+    {
+      "epoch": 3.976878612716763,
+      "grad_norm": 0.1307432598539227,
+      "learning_rate": 9.693181462235283e-06,
+      "loss": 0.4555,
+      "step": 1290
+    },
+    {
+      "epoch": 3.979961464354528,
+      "grad_norm": 0.11689497341729971,
+      "learning_rate": 9.637001638426456e-06,
+      "loss": 0.4567,
+      "step": 1291
+    },
+    {
+      "epoch": 3.983044315992293,
+      "grad_norm": 0.1252690386365334,
+      "learning_rate": 9.580962785424517e-06,
+      "loss": 0.4595,
+      "step": 1292
+    },
+    {
+      "epoch": 3.986127167630058,
+      "grad_norm": 0.1307010454041195,
+      "learning_rate": 9.525065163409363e-06,
+      "loss": 0.4572,
+      "step": 1293
+    },
+    {
+      "epoch": 3.989210019267823,
+      "grad_norm": 0.11694513696339244,
+      "learning_rate": 9.469309031905216e-06,
+      "loss": 0.4592,
+      "step": 1294
+    },
+    {
+      "epoch": 3.9922928709055876,
+      "grad_norm": 0.1259741426755351,
+      "learning_rate": 9.413694649779321e-06,
+      "loss": 0.4557,
+      "step": 1295
+    },
+    {
+      "epoch": 3.9953757225433524,
+      "grad_norm": 0.11667608639085636,
+      "learning_rate": 9.358222275240884e-06,
+      "loss": 0.4571,
+      "step": 1296
+    },
+    {
+      "epoch": 3.9984585741811176,
+      "grad_norm": 0.11552935763778427,
+      "learning_rate": 9.302892165839713e-06,
+      "loss": 0.4542,
+      "step": 1297
+    },
+    {
+      "epoch": 4.001541425818883,
+      "grad_norm": 0.22721298677866622,
+      "learning_rate": 9.247704578465165e-06,
+      "loss": 0.7322,
+      "step": 1298
+    },
+    {
+      "epoch": 4.004624277456648,
+      "grad_norm": 0.14537949653131974,
+      "learning_rate": 9.19265976934482e-06,
+      "loss": 0.4346,
+      "step": 1299
+    },
+    {
+      "epoch": 4.007707129094412,
+      "grad_norm": 0.1523450848612795,
+      "learning_rate": 9.137757994043416e-06,
+      "loss": 0.4407,
+      "step": 1300
+    },
+    {
+      "epoch": 4.010789980732177,
+      "grad_norm": 0.16374365419230436,
+      "learning_rate": 9.082999507461587e-06,
+      "loss": 0.4367,
+      "step": 1301
+    },
+    {
+      "epoch": 4.013872832369942,
+      "grad_norm": 0.1385915982654427,
+      "learning_rate": 9.028384563834703e-06,
+      "loss": 0.4403,
+      "step": 1302
+    },
+    {
+      "epoch": 4.016955684007707,
+      "grad_norm": 0.16333098823303244,
+      "learning_rate": 8.973913416731688e-06,
+      "loss": 0.4394,
+      "step": 1303
+    },
+    {
+      "epoch": 4.020038535645472,
+      "grad_norm": 0.1529223655733566,
+      "learning_rate": 8.919586319053821e-06,
+      "loss": 0.439,
+      "step": 1304
+    },
+    {
+      "epoch": 4.023121387283237,
+      "grad_norm": 0.13524521030137274,
+      "learning_rate": 8.8654035230336e-06,
+      "loss": 0.4385,
+      "step": 1305
+    },
+    {
+      "epoch": 4.026204238921002,
+      "grad_norm": 0.16372008403719154,
+      "learning_rate": 8.81136528023355e-06,
+      "loss": 0.4406,
+      "step": 1306
+    },
+    {
+      "epoch": 4.029287090558767,
+      "grad_norm": 0.1587084288871606,
+      "learning_rate": 8.75747184154507e-06,
+      "loss": 0.4366,
+      "step": 1307
+    },
+    {
+      "epoch": 4.0323699421965316,
+      "grad_norm": 0.1351223184878644,
+      "learning_rate": 8.70372345718721e-06,
+      "loss": 0.4415,
+      "step": 1308
+    },
+    {
+      "epoch": 4.035452793834296,
+      "grad_norm": 0.15510410982628026,
+      "learning_rate": 8.65012037670562e-06,
+      "loss": 0.4385,
+      "step": 1309
+    },
+    {
+      "epoch": 4.038535645472062,
+      "grad_norm": 0.131901823854703,
+      "learning_rate": 8.596662848971245e-06,
+      "loss": 0.4396,
+      "step": 1310
+    },
+    {
+      "epoch": 4.041618497109827,
+      "grad_norm": 0.12903795731564358,
+      "learning_rate": 8.543351122179352e-06,
+      "loss": 0.4351,
+      "step": 1311
+    },
+    {
+      "epoch": 4.044701348747592,
+      "grad_norm": 0.1281716724708245,
+      "learning_rate": 8.490185443848183e-06,
+      "loss": 0.4421,
+      "step": 1312
+    },
+    {
+      "epoch": 4.047784200385356,
+      "grad_norm": 0.13543724129330748,
+      "learning_rate": 8.43716606081796e-06,
+      "loss": 0.4363,
+      "step": 1313
+    },
+    {
+      "epoch": 4.050867052023121,
+      "grad_norm": 0.12293692776859114,
+      "learning_rate": 8.384293219249633e-06,
+      "loss": 0.4374,
+      "step": 1314
+    },
+    {
+      "epoch": 4.053949903660886,
+      "grad_norm": 0.11789890526928858,
+      "learning_rate": 8.331567164623818e-06,
+      "loss": 0.433,
+      "step": 1315
+    },
+    {
+      "epoch": 4.057032755298652,
+      "grad_norm": 0.12903520934000112,
+      "learning_rate": 8.278988141739605e-06,
+      "loss": 0.4378,
+      "step": 1316
+    },
+    {
+      "epoch": 4.060115606936416,
+      "grad_norm": 0.12579555934302328,
+      "learning_rate": 8.226556394713459e-06,
+      "loss": 0.4353,
+      "step": 1317
+    },
+    {
+      "epoch": 4.063198458574181,
+      "grad_norm": 0.11215605862286049,
+      "learning_rate": 8.174272166978015e-06,
+      "loss": 0.4394,
+      "step": 1318
+    },
+    {
+      "epoch": 4.066281310211946,
+      "grad_norm": 0.11478799104985912,
+      "learning_rate": 8.122135701281051e-06,
+      "loss": 0.4393,
+      "step": 1319
+    },
+    {
+      "epoch": 4.069364161849711,
+      "grad_norm": 0.1295168577656691,
+      "learning_rate": 8.07014723968428e-06,
+      "loss": 0.4377,
+      "step": 1320
+    },
+    {
+      "epoch": 4.0724470134874755,
+      "grad_norm": 0.11907161089906149,
+      "learning_rate": 8.018307023562281e-06,
+      "loss": 0.4416,
+      "step": 1321
+    },
+    {
+      "epoch": 4.075529865125241,
+      "grad_norm": 0.11350493816009986,
+      "learning_rate": 7.966615293601308e-06,
+      "loss": 0.4384,
+      "step": 1322
+    },
+    {
+      "epoch": 4.078612716763006,
+      "grad_norm": 0.11853395657688186,
+      "learning_rate": 7.915072289798247e-06,
+      "loss": 0.4376,
+      "step": 1323
+    },
+    {
+      "epoch": 4.081695568400771,
+      "grad_norm": 0.11788562741176781,
+      "learning_rate": 7.863678251459483e-06,
+      "loss": 0.4389,
+      "step": 1324
+    },
+    {
+      "epoch": 4.0847784200385355,
+      "grad_norm": 0.11986206346272227,
+      "learning_rate": 7.812433417199727e-06,
+      "loss": 0.4383,
+      "step": 1325
+    },
+    {
+      "epoch": 4.0878612716763,
+      "grad_norm": 0.11377906021876016,
+      "learning_rate": 7.76133802494103e-06,
+      "loss": 0.4335,
+      "step": 1326
+    },
+    {
+      "epoch": 4.090944123314065,
+      "grad_norm": 0.11995348411451266,
+      "learning_rate": 7.710392311911534e-06,
+      "loss": 0.4405,
+      "step": 1327
+    },
+    {
+      "epoch": 4.094026974951831,
+      "grad_norm": 0.11654925054415054,
+      "learning_rate": 7.659596514644496e-06,
+      "loss": 0.4351,
+      "step": 1328
+    },
+    {
+      "epoch": 4.0971098265895955,
+      "grad_norm": 0.12032119141679685,
+      "learning_rate": 7.608950868977101e-06,
+      "loss": 0.4382,
+      "step": 1329
+    },
+    {
+      "epoch": 4.10019267822736,
+      "grad_norm": 0.11215116855696312,
+      "learning_rate": 7.5584556100494196e-06,
+      "loss": 0.4392,
+      "step": 1330
+    },
+    {
+      "epoch": 4.103275529865125,
+      "grad_norm": 0.10901305258417135,
+      "learning_rate": 7.508110972303306e-06,
+      "loss": 0.4375,
+      "step": 1331
+    },
+    {
+      "epoch": 4.10635838150289,
+      "grad_norm": 0.11345772159277048,
+      "learning_rate": 7.457917189481301e-06,
+      "loss": 0.4372,
+      "step": 1332
+    },
+    {
+      "epoch": 4.109441233140655,
+      "grad_norm": 0.11598238007777345,
+      "learning_rate": 7.407874494625527e-06,
+      "loss": 0.4385,
+      "step": 1333
+    },
+    {
+      "epoch": 4.11252408477842,
+      "grad_norm": 0.11246746692789124,
+      "learning_rate": 7.357983120076659e-06,
+      "loss": 0.4381,
+      "step": 1334
+    },
+    {
+      "epoch": 4.115606936416185,
+      "grad_norm": 0.11495440965194466,
+      "learning_rate": 7.308243297472781e-06,
+      "loss": 0.4392,
+      "step": 1335
+    },
+    {
+      "epoch": 4.11868978805395,
+      "grad_norm": 0.10702555697060594,
+      "learning_rate": 7.2586552577483725e-06,
+      "loss": 0.4375,
+      "step": 1336
+    },
+    {
+      "epoch": 4.121772639691715,
+      "grad_norm": 0.10924475653202677,
+      "learning_rate": 7.209219231133202e-06,
+      "loss": 0.4389,
+      "step": 1337
+    },
+    {
+      "epoch": 4.1248554913294795,
+      "grad_norm": 0.12014291837981717,
+      "learning_rate": 7.159935447151265e-06,
+      "loss": 0.4387,
+      "step": 1338
+    },
+    {
+      "epoch": 4.127938342967244,
+      "grad_norm": 0.1058610470444985,
+      "learning_rate": 7.1108041346197e-06,
+      "loss": 0.4398,
+      "step": 1339
+    },
+    {
+      "epoch": 4.13102119460501,
+      "grad_norm": 0.10536763657511629,
+      "learning_rate": 7.061825521647758e-06,
+      "loss": 0.4391,
+      "step": 1340
+    },
+    {
+      "epoch": 4.134104046242775,
+      "grad_norm": 0.1079813591234523,
+      "learning_rate": 7.0129998356357295e-06,
+      "loss": 0.4386,
+      "step": 1341
+    },
+    {
+      "epoch": 4.1371868978805395,
+      "grad_norm": 0.1080418829328855,
+      "learning_rate": 6.96432730327389e-06,
+      "loss": 0.4407,
+      "step": 1342
+    },
+    {
+      "epoch": 4.140269749518304,
+      "grad_norm": 0.12097465696875707,
+      "learning_rate": 6.915808150541452e-06,
+      "loss": 0.4388,
+      "step": 1343
+    },
+    {
+      "epoch": 4.143352601156069,
+      "grad_norm": 0.09900314480713641,
+      "learning_rate": 6.867442602705478e-06,
+      "loss": 0.4419,
+      "step": 1344
+    },
+    {
+      "epoch": 4.146435452793835,
+      "grad_norm": 0.10170732242784322,
+      "learning_rate": 6.81923088431991e-06,
+      "loss": 0.4384,
+      "step": 1345
+    },
+    {
+      "epoch": 4.1495183044315995,
+      "grad_norm": 0.11176383898171684,
+      "learning_rate": 6.771173219224425e-06,
+      "loss": 0.4325,
+      "step": 1346
+    },
+    {
+      "epoch": 4.152601156069364,
+      "grad_norm": 0.10935988968219403,
+      "learning_rate": 6.72326983054354e-06,
+      "loss": 0.4375,
+      "step": 1347
+    },
+    {
+      "epoch": 4.155684007707129,
+      "grad_norm": 0.10269401213984333,
+      "learning_rate": 6.675520940685398e-06,
+      "loss": 0.4365,
+      "step": 1348
+    },
+    {
+      "epoch": 4.158766859344894,
+      "grad_norm": 0.10646286658255982,
+      "learning_rate": 6.627926771340902e-06,
+      "loss": 0.4386,
+      "step": 1349
+    },
+    {
+      "epoch": 4.161849710982659,
+      "grad_norm": 0.11654541514479556,
+      "learning_rate": 6.58048754348255e-06,
+      "loss": 0.4406,
+      "step": 1350
+    },
+    {
+      "epoch": 4.164932562620423,
+      "grad_norm": 0.10430033147328988,
+      "learning_rate": 6.533203477363508e-06,
+      "loss": 0.4355,
+      "step": 1351
+    },
+    {
+      "epoch": 4.168015414258189,
+      "grad_norm": 0.10202604264479437,
+      "learning_rate": 6.4860747925165415e-06,
+      "loss": 0.4367,
+      "step": 1352
+    },
+    {
+      "epoch": 4.171098265895954,
+      "grad_norm": 0.10303717616879111,
+      "learning_rate": 6.439101707753015e-06,
+      "loss": 0.4368,
+      "step": 1353
+    },
+    {
+      "epoch": 4.174181117533719,
+      "grad_norm": 0.09871062957252229,
+      "learning_rate": 6.392284441161835e-06,
+      "loss": 0.4367,
+      "step": 1354
+    },
+    {
+      "epoch": 4.1772639691714835,
+      "grad_norm": 0.10437326316193471,
+      "learning_rate": 6.345623210108498e-06,
+      "loss": 0.4391,
+      "step": 1355
+    },
+    {
+      "epoch": 4.180346820809248,
+      "grad_norm": 0.1114519572334138,
+      "learning_rate": 6.299118231234018e-06,
+      "loss": 0.4345,
+      "step": 1356
+    },
+    {
+      "epoch": 4.183429672447014,
+      "grad_norm": 0.10616244141924702,
+      "learning_rate": 6.252769720454006e-06,
+      "loss": 0.4365,
+      "step": 1357
+    },
+    {
+      "epoch": 4.186512524084779,
+      "grad_norm": 0.09943414781141079,
+      "learning_rate": 6.206577892957564e-06,
+      "loss": 0.4374,
+      "step": 1358
+    },
+    {
+      "epoch": 4.1895953757225435,
+      "grad_norm": 0.10282103345175542,
+      "learning_rate": 6.160542963206357e-06,
+      "loss": 0.4377,
+      "step": 1359
+    },
+    {
+      "epoch": 4.192678227360308,
+      "grad_norm": 0.10292193127599197,
+      "learning_rate": 6.114665144933614e-06,
+      "loss": 0.4368,
+      "step": 1360
+    },
+    {
+      "epoch": 4.195761078998073,
+      "grad_norm": 0.10228929645026591,
+      "learning_rate": 6.068944651143085e-06,
+      "loss": 0.4398,
+      "step": 1361
+    },
+    {
+      "epoch": 4.198843930635838,
+      "grad_norm": 0.10350507319151779,
+      "learning_rate": 6.023381694108108e-06,
+      "loss": 0.437,
+      "step": 1362
+    },
+    {
+      "epoch": 4.2019267822736035,
+      "grad_norm": 0.09822177264700199,
+      "learning_rate": 5.9779764853706e-06,
+      "loss": 0.4392,
+      "step": 1363
+    },
+    {
+      "epoch": 4.205009633911368,
+      "grad_norm": 0.10292863543420692,
+      "learning_rate": 5.932729235740087e-06,
+      "loss": 0.4364,
+      "step": 1364
+    },
+    {
+      "epoch": 4.208092485549133,
+      "grad_norm": 0.11060367900661935,
+      "learning_rate": 5.88764015529268e-06,
+      "loss": 0.4364,
+      "step": 1365
+    },
+    {
+      "epoch": 4.211175337186898,
+      "grad_norm": 0.10320184338784077,
+      "learning_rate": 5.842709453370172e-06,
+      "loss": 0.4376,
+      "step": 1366
+    },
+    {
+      "epoch": 4.214258188824663,
+      "grad_norm": 0.10329519268034482,
+      "learning_rate": 5.797937338579003e-06,
+      "loss": 0.4377,
+      "step": 1367
+    },
+    {
+      "epoch": 4.217341040462427,
+      "grad_norm": 0.10850415739692651,
+      "learning_rate": 5.753324018789346e-06,
+      "loss": 0.433,
+      "step": 1368
+    },
+    {
+      "epoch": 4.220423892100193,
+      "grad_norm": 0.09901495695931094,
+      "learning_rate": 5.708869701134067e-06,
+      "loss": 0.4388,
+      "step": 1369
+    },
+    {
+      "epoch": 4.223506743737958,
+      "grad_norm": 0.09975161582717798,
+      "learning_rate": 5.6645745920078606e-06,
+      "loss": 0.4358,
+      "step": 1370
+    },
+    {
+      "epoch": 4.226589595375723,
+      "grad_norm": 0.09796658952241978,
+      "learning_rate": 5.6204388970662e-06,
+      "loss": 0.438,
+      "step": 1371
+    },
+    {
+      "epoch": 4.229672447013487,
+      "grad_norm": 0.09762249144634376,
+      "learning_rate": 5.576462821224438e-06,
+      "loss": 0.4366,
+      "step": 1372
+    },
+    {
+      "epoch": 4.232755298651252,
+      "grad_norm": 0.09711078459256982,
+      "learning_rate": 5.532646568656845e-06,
+      "loss": 0.4403,
+      "step": 1373
+    },
+    {
+      "epoch": 4.235838150289017,
+      "grad_norm": 0.09423309298700806,
+      "learning_rate": 5.488990342795664e-06,
+      "loss": 0.4377,
+      "step": 1374
+    },
+    {
+      "epoch": 4.238921001926783,
+      "grad_norm": 0.10458221128175979,
+      "learning_rate": 5.445494346330122e-06,
+      "loss": 0.4382,
+      "step": 1375
+    },
+    {
+      "epoch": 4.2420038535645475,
+      "grad_norm": 0.10235466521290817,
+      "learning_rate": 5.402158781205562e-06,
+      "loss": 0.4359,
+      "step": 1376
+    },
+    {
+      "epoch": 4.245086705202312,
+      "grad_norm": 0.1058993604356189,
+      "learning_rate": 5.358983848622452e-06,
+      "loss": 0.4349,
+      "step": 1377
+    },
+    {
+      "epoch": 4.248169556840077,
+      "grad_norm": 0.10654331625195929,
+      "learning_rate": 5.3159697490354855e-06,
+      "loss": 0.4366,
+      "step": 1378
+    },
+    {
+      "epoch": 4.251252408477842,
+      "grad_norm": 0.09769527385848033,
+      "learning_rate": 5.273116682152593e-06,
+      "loss": 0.4349,
+      "step": 1379
+    },
+    {
+      "epoch": 4.254335260115607,
+      "grad_norm": 0.10715019266224053,
+      "learning_rate": 5.230424846934088e-06,
+      "loss": 0.4423,
+      "step": 1380
+    },
+    {
+      "epoch": 4.257418111753372,
+      "grad_norm": 0.10933113122312244,
+      "learning_rate": 5.1878944415917164e-06,
+      "loss": 0.4384,
+      "step": 1381
+    },
+    {
+      "epoch": 4.260500963391137,
+      "grad_norm": 0.10741363531965316,
+      "learning_rate": 5.145525663587685e-06,
+      "loss": 0.4325,
+      "step": 1382
+    },
+    {
+      "epoch": 4.263583815028902,
+      "grad_norm": 0.10325059146559737,
+      "learning_rate": 5.103318709633853e-06,
+      "loss": 0.4385,
+      "step": 1383
+    },
+    {
+      "epoch": 4.266666666666667,
+      "grad_norm": 0.10781380541904591,
+      "learning_rate": 5.061273775690701e-06,
+      "loss": 0.4377,
+      "step": 1384
+    },
+    {
+      "epoch": 4.269749518304431,
+      "grad_norm": 0.10954484488198808,
+      "learning_rate": 5.019391056966503e-06,
+      "loss": 0.4341,
+      "step": 1385
+    },
+    {
+      "epoch": 4.272832369942196,
+      "grad_norm": 0.10301443923623921,
+      "learning_rate": 4.97767074791637e-06,
+      "loss": 0.4353,
+      "step": 1386
+    },
+    {
+      "epoch": 4.275915221579962,
+      "grad_norm": 0.1009381174729872,
+      "learning_rate": 4.936113042241397e-06,
+      "loss": 0.4414,
+      "step": 1387
+    },
+    {
+      "epoch": 4.278998073217727,
+      "grad_norm": 0.10335307452114761,
+      "learning_rate": 4.894718132887714e-06,
+      "loss": 0.4381,
+      "step": 1388
+    },
+    {
+      "epoch": 4.282080924855491,
+      "grad_norm": 0.10115315083359568,
+      "learning_rate": 4.853486212045644e-06,
+      "loss": 0.436,
+      "step": 1389
+    },
+    {
+      "epoch": 4.285163776493256,
+      "grad_norm": 0.10876623934591757,
+      "learning_rate": 4.812417471148725e-06,
+      "loss": 0.4342,
+      "step": 1390
+    },
+    {
+      "epoch": 4.288246628131021,
+      "grad_norm": 0.09704299284424527,
+      "learning_rate": 4.77151210087293e-06,
+      "loss": 0.4364,
+      "step": 1391
+    },
+    {
+      "epoch": 4.291329479768786,
+      "grad_norm": 0.10292000396224527,
+      "learning_rate": 4.73077029113568e-06,
+      "loss": 0.434,
+      "step": 1392
+    },
+    {
+      "epoch": 4.294412331406551,
+      "grad_norm": 0.10155539343117652,
+      "learning_rate": 4.690192231095063e-06,
+      "loss": 0.4354,
+      "step": 1393
+    },
+    {
+      "epoch": 4.297495183044316,
+      "grad_norm": 0.10055822887935521,
+      "learning_rate": 4.649778109148844e-06,
+      "loss": 0.4387,
+      "step": 1394
+    },
+    {
+      "epoch": 4.300578034682081,
+      "grad_norm": 0.09837343573960855,
+      "learning_rate": 4.609528112933688e-06,
+      "loss": 0.4335,
+      "step": 1395
+    },
+    {
+      "epoch": 4.303660886319846,
+      "grad_norm": 0.10376560621484232,
+      "learning_rate": 4.5694424293242445e-06,
+      "loss": 0.4394,
+      "step": 1396
+    },
+    {
+      "epoch": 4.306743737957611,
+      "grad_norm": 0.099601803564409,
+      "learning_rate": 4.529521244432262e-06,
+      "loss": 0.4365,
+      "step": 1397
+    },
+    {
+      "epoch": 4.309826589595375,
+      "grad_norm": 0.102945557544895,
+      "learning_rate": 4.489764743605767e-06,
+      "loss": 0.4364,
+      "step": 1398
+    },
+    {
+      "epoch": 4.312909441233141,
+      "grad_norm": 0.09732849693536313,
+      "learning_rate": 4.450173111428173e-06,
+      "loss": 0.4358,
+      "step": 1399
+    },
+    {
+      "epoch": 4.315992292870906,
+      "grad_norm": 0.08996607783762468,
+      "learning_rate": 4.410746531717452e-06,
+      "loss": 0.4387,
+      "step": 1400
+    },
+    {
+      "epoch": 4.319075144508671,
+      "grad_norm": 0.09888296412031398,
+      "learning_rate": 4.371485187525228e-06,
+      "loss": 0.4367,
+      "step": 1401
+    },
+    {
+      "epoch": 4.322157996146435,
+      "grad_norm": 0.09644944594307794,
+      "learning_rate": 4.332389261135994e-06,
+      "loss": 0.4404,
+      "step": 1402
+    },
+    {
+      "epoch": 4.3252408477842,
+      "grad_norm": 0.11172318268123643,
+      "learning_rate": 4.2934589340662125e-06,
+      "loss": 0.4416,
+      "step": 1403
+    },
+    {
+      "epoch": 4.328323699421965,
+      "grad_norm": 0.09624635988379959,
+      "learning_rate": 4.254694387063514e-06,
+      "loss": 0.4355,
+      "step": 1404
+    },
+    {
+      "epoch": 4.331406551059731,
+      "grad_norm": 0.09608363334298854,
+      "learning_rate": 4.2160958001058065e-06,
+      "loss": 0.4375,
+      "step": 1405
+    },
+    {
+      "epoch": 4.334489402697495,
+      "grad_norm": 0.10129838608553117,
+      "learning_rate": 4.1776633524005075e-06,
+      "loss": 0.4385,
+      "step": 1406
+    },
+    {
+      "epoch": 4.33757225433526,
+      "grad_norm": 0.09298789697076024,
+      "learning_rate": 4.1393972223836255e-06,
+      "loss": 0.4365,
+      "step": 1407
+    },
+    {
+      "epoch": 4.340655105973025,
+      "grad_norm": 0.09333757526368255,
+      "learning_rate": 4.101297587719053e-06,
+      "loss": 0.4392,
+      "step": 1408
+    },
+    {
+      "epoch": 4.34373795761079,
+      "grad_norm": 0.09376559970193508,
+      "learning_rate": 4.063364625297599e-06,
+      "loss": 0.4366,
+      "step": 1409
+    },
+    {
+      "epoch": 4.3468208092485545,
+      "grad_norm": 0.09287130745095976,
+      "learning_rate": 4.025598511236282e-06,
+      "loss": 0.4371,
+      "step": 1410
+    },
+    {
+      "epoch": 4.34990366088632,
+      "grad_norm": 0.09559999506644899,
+      "learning_rate": 3.987999420877442e-06,
+      "loss": 0.4389,
+      "step": 1411
+    },
+    {
+      "epoch": 4.352986512524085,
+      "grad_norm": 0.09943125985290735,
+      "learning_rate": 3.9505675287879785e-06,
+      "loss": 0.4385,
+      "step": 1412
+    },
+    {
+      "epoch": 4.35606936416185,
+      "grad_norm": 0.08868252059738668,
+      "learning_rate": 3.913303008758491e-06,
+      "loss": 0.4392,
+      "step": 1413
+    },
+    {
+      "epoch": 4.3591522157996145,
+      "grad_norm": 0.09515486848275258,
+      "learning_rate": 3.876206033802525e-06,
+      "loss": 0.4339,
+      "step": 1414
+    },
+    {
+      "epoch": 4.362235067437379,
+      "grad_norm": 0.0960065547465076,
+      "learning_rate": 3.839276776155703e-06,
+      "loss": 0.4381,
+      "step": 1415
+    },
+    {
+      "epoch": 4.365317919075144,
+      "grad_norm": 0.09040026109168876,
+      "learning_rate": 3.8025154072749872e-06,
+      "loss": 0.4355,
+      "step": 1416
+    },
+    {
+      "epoch": 4.36840077071291,
+      "grad_norm": 0.102041325954044,
+      "learning_rate": 3.765922097837855e-06,
+      "loss": 0.4357,
+      "step": 1417
+    },
+    {
+      "epoch": 4.371483622350675,
+      "grad_norm": 0.10036694707001045,
+      "learning_rate": 3.7294970177414923e-06,
+      "loss": 0.437,
+      "step": 1418
+    },
+    {
+      "epoch": 4.374566473988439,
+      "grad_norm": 0.09612050182063445,
+      "learning_rate": 3.693240336102051e-06,
+      "loss": 0.4392,
+      "step": 1419
+    },
+    {
+      "epoch": 4.377649325626204,
+      "grad_norm": 0.09305353518081604,
+      "learning_rate": 3.657152221253806e-06,
+      "loss": 0.437,
+      "step": 1420
+    },
+    {
+      "epoch": 4.380732177263969,
+      "grad_norm": 0.09137003873507461,
+      "learning_rate": 3.6212328407484233e-06,
+      "loss": 0.4365,
+      "step": 1421
+    },
+    {
+      "epoch": 4.383815028901734,
+      "grad_norm": 0.09815952383999023,
+      "learning_rate": 3.585482361354138e-06,
+      "loss": 0.4387,
+      "step": 1422
+    },
+    {
+      "epoch": 4.386897880539499,
+      "grad_norm": 0.09725005309124897,
+      "learning_rate": 3.5499009490550208e-06,
+      "loss": 0.4372,
+      "step": 1423
+    },
+    {
+      "epoch": 4.389980732177264,
+      "grad_norm": 0.0942493336766283,
+      "learning_rate": 3.514488769050188e-06,
+      "loss": 0.4388,
+      "step": 1424
+    },
+    {
+      "epoch": 4.393063583815029,
+      "grad_norm": 0.09545240417898013,
+      "learning_rate": 3.4792459857530393e-06,
+      "loss": 0.4373,
+      "step": 1425
+    },
+    {
+      "epoch": 4.396146435452794,
+      "grad_norm": 0.09520241207307875,
+      "learning_rate": 3.444172762790472e-06,
+      "loss": 0.4361,
+      "step": 1426
+    },
+    {
+      "epoch": 4.3992292870905585,
+      "grad_norm": 0.0885822279530784,
+      "learning_rate": 3.4092692630021617e-06,
+      "loss": 0.438,
+      "step": 1427
+    },
+    {
+      "epoch": 4.402312138728323,
+      "grad_norm": 0.09853294252317715,
+      "learning_rate": 3.3745356484397653e-06,
+      "loss": 0.4391,
+      "step": 1428
+    },
+    {
+      "epoch": 4.405394990366089,
+      "grad_norm": 0.10391907601726293,
+      "learning_rate": 3.339972080366214e-06,
+      "loss": 0.4402,
+      "step": 1429
+    },
+    {
+      "epoch": 4.408477842003854,
+      "grad_norm": 0.09550139048249519,
+      "learning_rate": 3.305578719254916e-06,
+      "loss": 0.4374,
+      "step": 1430
+    },
+    {
+      "epoch": 4.4115606936416185,
+      "grad_norm": 0.09451462600644592,
+      "learning_rate": 3.2713557247890447e-06,
+      "loss": 0.4375,
+      "step": 1431
+    },
+    {
+      "epoch": 4.414643545279383,
+      "grad_norm": 0.09286873347182731,
+      "learning_rate": 3.23730325586078e-06,
+      "loss": 0.44,
+      "step": 1432
+    },
+    {
+      "epoch": 4.417726396917148,
+      "grad_norm": 0.10123180464832394,
+      "learning_rate": 3.2034214705705867e-06,
+      "loss": 0.4419,
+      "step": 1433
+    },
+    {
+      "epoch": 4.420809248554914,
+      "grad_norm": 0.10285255781475613,
+      "learning_rate": 3.1697105262264684e-06,
+      "loss": 0.4374,
+      "step": 1434
+    },
+    {
+      "epoch": 4.4238921001926785,
+      "grad_norm": 0.09682264930518235,
+      "learning_rate": 3.136170579343234e-06,
+      "loss": 0.4353,
+      "step": 1435
+    },
+    {
+      "epoch": 4.426974951830443,
+      "grad_norm": 0.09709235344577144,
+      "learning_rate": 3.1028017856417916e-06,
+      "loss": 0.4374,
+      "step": 1436
+    },
+    {
+      "epoch": 4.430057803468208,
+      "grad_norm": 0.08767204735403385,
+      "learning_rate": 3.0696043000483856e-06,
+      "loss": 0.438,
+      "step": 1437
+    },
+    {
+      "epoch": 4.433140655105973,
+      "grad_norm": 0.0932966051033176,
+      "learning_rate": 3.0365782766939244e-06,
+      "loss": 0.439,
+      "step": 1438
+    },
+    {
+      "epoch": 4.436223506743738,
+      "grad_norm": 0.09829842354923603,
+      "learning_rate": 3.003723868913242e-06,
+      "loss": 0.4372,
+      "step": 1439
+    },
+    {
+      "epoch": 4.4393063583815024,
+      "grad_norm": 0.104432475355302,
+      "learning_rate": 2.9710412292443868e-06,
+      "loss": 0.4349,
+      "step": 1440
+    },
+    {
+      "epoch": 4.442389210019268,
+      "grad_norm": 0.09047991694536207,
+      "learning_rate": 2.9385305094278947e-06,
+      "loss": 0.4349,
+      "step": 1441
+    },
+    {
+      "epoch": 4.445472061657033,
+      "grad_norm": 0.0966893975084566,
+      "learning_rate": 2.906191860406131e-06,
+      "loss": 0.4425,
+      "step": 1442
+    },
+    {
+      "epoch": 4.448554913294798,
+      "grad_norm": 0.09149963030504712,
+      "learning_rate": 2.874025432322527e-06,
+      "loss": 0.439,
+      "step": 1443
+    },
+    {
+      "epoch": 4.4516377649325625,
+      "grad_norm": 0.09446384441642167,
+      "learning_rate": 2.842031374520975e-06,
+      "loss": 0.4408,
+      "step": 1444
+    },
+    {
+      "epoch": 4.454720616570327,
+      "grad_norm": 0.10425191772432348,
+      "learning_rate": 2.810209835545017e-06,
+      "loss": 0.4359,
+      "step": 1445
+    },
+    {
+      "epoch": 4.457803468208093,
+      "grad_norm": 0.0961686350162677,
+      "learning_rate": 2.7785609631372624e-06,
+      "loss": 0.4409,
+      "step": 1446
+    },
+    {
+      "epoch": 4.460886319845858,
+      "grad_norm": 0.0896372031272348,
+      "learning_rate": 2.7470849042386196e-06,
+      "loss": 0.4413,
+      "step": 1447
+    },
+    {
+      "epoch": 4.4639691714836225,
+      "grad_norm": 0.0918031066655652,
+      "learning_rate": 2.715781804987674e-06,
+      "loss": 0.4424,
+      "step": 1448
+    },
+    {
+      "epoch": 4.467052023121387,
+      "grad_norm": 0.09476040432622543,
+      "learning_rate": 2.6846518107199782e-06,
+      "loss": 0.4366,
+      "step": 1449
+    },
+    {
+      "epoch": 4.470134874759152,
+      "grad_norm": 0.09676658586694219,
+      "learning_rate": 2.6536950659673945e-06,
+      "loss": 0.4385,
+      "step": 1450
+    },
+    {
+      "epoch": 4.473217726396917,
+      "grad_norm": 0.10207623891967303,
+      "learning_rate": 2.622911714457388e-06,
+      "loss": 0.4355,
+      "step": 1451
+    },
+    {
+      "epoch": 4.476300578034682,
+      "grad_norm": 0.0914148134007726,
+      "learning_rate": 2.5923018991124148e-06,
+      "loss": 0.4401,
+      "step": 1452
+    },
+    {
+      "epoch": 4.479383429672447,
+      "grad_norm": 0.08880708292877328,
+      "learning_rate": 2.5618657620492114e-06,
+      "loss": 0.4422,
+      "step": 1453
+    },
+    {
+      "epoch": 4.482466281310212,
+      "grad_norm": 0.09675309674244707,
+      "learning_rate": 2.53160344457815e-06,
+      "loss": 0.4354,
+      "step": 1454
+    },
+    {
+      "epoch": 4.485549132947977,
+      "grad_norm": 0.0987168013643815,
+      "learning_rate": 2.5015150872025996e-06,
+      "loss": 0.4373,
+      "step": 1455
+    },
+    {
+      "epoch": 4.488631984585742,
+      "grad_norm": 0.10151772653303257,
+      "learning_rate": 2.471600829618246e-06,
+      "loss": 0.4303,
+      "step": 1456
+    },
+    {
+      "epoch": 4.491714836223506,
+      "grad_norm": 0.09226750254863196,
+      "learning_rate": 2.441860810712471e-06,
+      "loss": 0.4445,
+      "step": 1457
+    },
+    {
+      "epoch": 4.494797687861272,
+      "grad_norm": 0.09165235630962736,
+      "learning_rate": 2.4122951685636674e-06,
+      "loss": 0.4385,
+      "step": 1458
+    },
+    {
+      "epoch": 4.497880539499037,
+      "grad_norm": 0.09311512907445026,
+      "learning_rate": 2.3829040404406413e-06,
+      "loss": 0.4395,
+      "step": 1459
+    },
+    {
+      "epoch": 4.500963391136802,
+      "grad_norm": 0.09374966399032056,
+      "learning_rate": 2.3536875628019606e-06,
+      "loss": 0.4374,
+      "step": 1460
+    },
+    {
+      "epoch": 4.5040462427745664,
+      "grad_norm": 0.10348130602014464,
+      "learning_rate": 2.32464587129531e-06,
+      "loss": 0.4365,
+      "step": 1461
+    },
+    {
+      "epoch": 4.507129094412331,
+      "grad_norm": 0.09988534801023742,
+      "learning_rate": 2.295779100756854e-06,
+      "loss": 0.4364,
+      "step": 1462
+    },
+    {
+      "epoch": 4.510211946050096,
+      "grad_norm": 0.0962822531671393,
+      "learning_rate": 2.2670873852106602e-06,
+      "loss": 0.438,
+      "step": 1463
+    },
+    {
+      "epoch": 4.513294797687861,
+      "grad_norm": 0.09628050923857297,
+      "learning_rate": 2.238570857868001e-06,
+      "loss": 0.4352,
+      "step": 1464
+    },
+    {
+      "epoch": 4.5163776493256265,
+      "grad_norm": 0.09600622582075857,
+      "learning_rate": 2.2102296511268183e-06,
+      "loss": 0.4376,
+      "step": 1465
+    },
+    {
+      "epoch": 4.519460500963391,
+      "grad_norm": 0.09282766627263783,
+      "learning_rate": 2.1820638965710295e-06,
+      "loss": 0.4374,
+      "step": 1466
+    },
+    {
+      "epoch": 4.522543352601156,
+      "grad_norm": 0.0973992257554959,
+      "learning_rate": 2.1540737249699893e-06,
+      "loss": 0.4347,
+      "step": 1467
+    },
+    {
+      "epoch": 4.525626204238921,
+      "grad_norm": 0.09522277402667699,
+      "learning_rate": 2.1262592662778083e-06,
+      "loss": 0.4395,
+      "step": 1468
+    },
+    {
+      "epoch": 4.528709055876686,
+      "grad_norm": 0.09144680443278094,
+      "learning_rate": 2.098620649632821e-06,
+      "loss": 0.4403,
+      "step": 1469
+    },
+    {
+      "epoch": 4.531791907514451,
+      "grad_norm": 0.08781416534415752,
+      "learning_rate": 2.0711580033569414e-06,
+      "loss": 0.4377,
+      "step": 1470
+    },
+    {
+      "epoch": 4.534874759152216,
+      "grad_norm": 0.09386810526841069,
+      "learning_rate": 2.0438714549550775e-06,
+      "loss": 0.4355,
+      "step": 1471
+    },
+    {
+      "epoch": 4.537957610789981,
+      "grad_norm": 0.09382659989627258,
+      "learning_rate": 2.0167611311145528e-06,
+      "loss": 0.4402,
+      "step": 1472
+    },
+    {
+      "epoch": 4.541040462427746,
+      "grad_norm": 0.09798605020490526,
+      "learning_rate": 1.989827157704487e-06,
+      "loss": 0.4351,
+      "step": 1473
+    },
+    {
+      "epoch": 4.54412331406551,
+      "grad_norm": 0.09116851181125603,
+      "learning_rate": 1.9630696597752495e-06,
+      "loss": 0.44,
+      "step": 1474
+    },
+    {
+      "epoch": 4.547206165703275,
+      "grad_norm": 0.09061635674266462,
+      "learning_rate": 1.936488761557853e-06,
+      "loss": 0.4348,
+      "step": 1475
+    },
+    {
+      "epoch": 4.55028901734104,
+      "grad_norm": 0.08822473493571764,
+      "learning_rate": 1.9100845864633875e-06,
+      "loss": 0.4369,
+      "step": 1476
+    },
+    {
+      "epoch": 4.553371868978806,
+      "grad_norm": 0.08881397137002002,
+      "learning_rate": 1.8838572570824397e-06,
+      "loss": 0.4393,
+      "step": 1477
+    },
+    {
+      "epoch": 4.55645472061657,
+      "grad_norm": 0.0902182458999888,
+      "learning_rate": 1.8578068951845373e-06,
+      "loss": 0.4374,
+      "step": 1478
+    },
+    {
+      "epoch": 4.559537572254335,
+      "grad_norm": 0.09001669107377033,
+      "learning_rate": 1.8319336217175543e-06,
+      "loss": 0.4411,
+      "step": 1479
+    },
+    {
+      "epoch": 4.5626204238921,
+      "grad_norm": 0.09363413452276474,
+      "learning_rate": 1.8062375568072e-06,
+      "loss": 0.4372,
+      "step": 1480
+    },
+    {
+      "epoch": 4.565703275529865,
+      "grad_norm": 0.08872466515536319,
+      "learning_rate": 1.7807188197564063e-06,
+      "loss": 0.438,
+      "step": 1481
+    },
+    {
+      "epoch": 4.56878612716763,
+      "grad_norm": 0.08846849680715307,
+      "learning_rate": 1.7553775290448172e-06,
+      "loss": 0.4385,
+      "step": 1482
+    },
+    {
+      "epoch": 4.571868978805395,
+      "grad_norm": 0.08880421567918338,
+      "learning_rate": 1.7302138023281978e-06,
+      "loss": 0.4322,
+      "step": 1483
+    },
+    {
+      "epoch": 4.57495183044316,
+      "grad_norm": 0.08723278420201072,
+      "learning_rate": 1.7052277564379282e-06,
+      "loss": 0.4358,
+      "step": 1484
+    },
+    {
+      "epoch": 4.578034682080925,
+      "grad_norm": 0.09032806954755906,
+      "learning_rate": 1.6804195073804442e-06,
+      "loss": 0.4408,
+      "step": 1485
+    },
+    {
+      "epoch": 4.58111753371869,
+      "grad_norm": 0.09791189876034624,
+      "learning_rate": 1.65578917033669e-06,
+      "loss": 0.4429,
+      "step": 1486
+    },
+    {
+      "epoch": 4.584200385356454,
+      "grad_norm": 0.08641173081535548,
+      "learning_rate": 1.631336859661592e-06,
+      "loss": 0.4382,
+      "step": 1487
+    },
+    {
+      "epoch": 4.58728323699422,
+      "grad_norm": 0.09020253360408878,
+      "learning_rate": 1.6070626888835227e-06,
+      "loss": 0.4343,
+      "step": 1488
+    },
+    {
+      "epoch": 4.590366088631985,
+      "grad_norm": 0.08675695308240004,
+      "learning_rate": 1.5829667707037976e-06,
+      "loss": 0.4365,
+      "step": 1489
+    },
+    {
+      "epoch": 4.59344894026975,
+      "grad_norm": 0.08705186692168186,
+      "learning_rate": 1.5590492169961046e-06,
+      "loss": 0.4431,
+      "step": 1490
+    },
+    {
+      "epoch": 4.596531791907514,
+      "grad_norm": 0.08700465591561538,
+      "learning_rate": 1.5353101388060343e-06,
+      "loss": 0.4341,
+      "step": 1491
+    },
+    {
+      "epoch": 4.599614643545279,
+      "grad_norm": 0.09068913252992127,
+      "learning_rate": 1.5117496463505333e-06,
+      "loss": 0.4318,
+      "step": 1492
+    },
+    {
+      "epoch": 4.602697495183044,
+      "grad_norm": 0.08232738526504484,
+      "learning_rate": 1.4883678490174025e-06,
+      "loss": 0.438,
+      "step": 1493
+    },
+    {
+      "epoch": 4.60578034682081,
+      "grad_norm": 0.08769111572205408,
+      "learning_rate": 1.4651648553647869e-06,
+      "loss": 0.4355,
+      "step": 1494
+    },
+    {
+      "epoch": 4.608863198458574,
+      "grad_norm": 0.0916927101757694,
+      "learning_rate": 1.4421407731206771e-06,
+      "loss": 0.4372,
+      "step": 1495
+    },
+    {
+      "epoch": 4.611946050096339,
+      "grad_norm": 0.09028398647083151,
+      "learning_rate": 1.4192957091824044e-06,
+      "loss": 0.4356,
+      "step": 1496
+    },
+    {
+      "epoch": 4.615028901734104,
+      "grad_norm": 0.09261226330584073,
+      "learning_rate": 1.3966297696161423e-06,
+      "loss": 0.4408,
+      "step": 1497
+    },
+    {
+      "epoch": 4.618111753371869,
+      "grad_norm": 0.08693079019511248,
+      "learning_rate": 1.3741430596564142e-06,
+      "loss": 0.4381,
+      "step": 1498
+    },
+    {
+      "epoch": 4.6211946050096335,
+      "grad_norm": 0.08614796846664992,
+      "learning_rate": 1.3518356837056224e-06,
+      "loss": 0.4376,
+      "step": 1499
+    },
+    {
+      "epoch": 4.624277456647399,
+      "grad_norm": 0.09063265333949524,
+      "learning_rate": 1.3297077453335194e-06,
+      "loss": 0.4356,
+      "step": 1500
+    },
+    {
+      "epoch": 4.627360308285164,
+      "grad_norm": 0.08762485167757508,
+      "learning_rate": 1.3077593472767958e-06,
+      "loss": 0.4389,
+      "step": 1501
+    },
+    {
+      "epoch": 4.630443159922929,
+      "grad_norm": 0.09006117033448659,
+      "learning_rate": 1.2859905914385283e-06,
+      "loss": 0.435,
+      "step": 1502
+    },
+    {
+      "epoch": 4.6335260115606935,
+      "grad_norm": 0.08399200079038739,
+      "learning_rate": 1.2644015788877684e-06,
+      "loss": 0.4363,
+      "step": 1503
+    },
+    {
+      "epoch": 4.636608863198458,
+      "grad_norm": 0.0902502698196518,
+      "learning_rate": 1.2429924098590207e-06,
+      "loss": 0.4393,
+      "step": 1504
+    },
+    {
+      "epoch": 4.639691714836223,
+      "grad_norm": 0.08372372123786889,
+      "learning_rate": 1.221763183751823e-06,
+      "loss": 0.4358,
+      "step": 1505
+    },
+    {
+      "epoch": 4.642774566473989,
+      "grad_norm": 0.08663989373549699,
+      "learning_rate": 1.2007139991302608e-06,
+      "loss": 0.4395,
+      "step": 1506
+    },
+    {
+      "epoch": 4.645857418111754,
+      "grad_norm": 0.08615090496613724,
+      "learning_rate": 1.1798449537225198e-06,
+      "loss": 0.4378,
+      "step": 1507
+    },
+    {
+      "epoch": 4.648940269749518,
+      "grad_norm": 0.08685465408923697,
+      "learning_rate": 1.1591561444204057e-06,
+      "loss": 0.4386,
+      "step": 1508
+    },
+    {
+      "epoch": 4.652023121387283,
+      "grad_norm": 0.08492672708997942,
+      "learning_rate": 1.1386476672789404e-06,
+      "loss": 0.4345,
+      "step": 1509
+    },
+    {
+      "epoch": 4.655105973025048,
+      "grad_norm": 0.08994269445781145,
+      "learning_rate": 1.1183196175158728e-06,
+      "loss": 0.4381,
+      "step": 1510
+    },
+    {
+      "epoch": 4.658188824662814,
+      "grad_norm": 0.08700948777373317,
+      "learning_rate": 1.0981720895112668e-06,
+      "loss": 0.432,
+      "step": 1511
+    },
+    {
+      "epoch": 4.661271676300578,
+      "grad_norm": 0.08702757811060335,
+      "learning_rate": 1.0782051768070477e-06,
+      "loss": 0.436,
+      "step": 1512
+    },
+    {
+      "epoch": 4.664354527938343,
+      "grad_norm": 0.086614725547039,
+      "learning_rate": 1.058418972106563e-06,
+      "loss": 0.4358,
+      "step": 1513
+    },
+    {
+      "epoch": 4.667437379576108,
+      "grad_norm": 0.08221794558096522,
+      "learning_rate": 1.0388135672741773e-06,
+      "loss": 0.4376,
+      "step": 1514
+    },
+    {
+      "epoch": 4.670520231213873,
+      "grad_norm": 0.0851154298657411,
+      "learning_rate": 1.0193890533348028e-06,
+      "loss": 0.436,
+      "step": 1515
+    },
+    {
+      "epoch": 4.6736030828516375,
+      "grad_norm": 0.08682560935634347,
+      "learning_rate": 1.0001455204735343e-06,
+      "loss": 0.4374,
+      "step": 1516
+    },
+    {
+      "epoch": 4.676685934489402,
+      "grad_norm": 0.09338827574575734,
+      "learning_rate": 9.81083058035175e-07,
+      "loss": 0.4394,
+      "step": 1517
+    },
+    {
+      "epoch": 4.679768786127168,
+      "grad_norm": 0.0830058269593175,
+      "learning_rate": 9.622017545238571e-07,
+      "loss": 0.436,
+      "step": 1518
+    },
+    {
+      "epoch": 4.682851637764933,
+      "grad_norm": 0.08302323854075715,
+      "learning_rate": 9.435016976026135e-07,
+      "loss": 0.4351,
+      "step": 1519
+    },
+    {
+      "epoch": 4.6859344894026975,
+      "grad_norm": 0.08637574002915817,
+      "learning_rate": 9.249829740929761e-07,
+      "loss": 0.4382,
+      "step": 1520
+    },
+    {
+      "epoch": 4.689017341040462,
+      "grad_norm": 0.08181061730485034,
+      "learning_rate": 9.066456699745774e-07,
+      "loss": 0.4325,
+      "step": 1521
+    },
+    {
+      "epoch": 4.692100192678227,
+      "grad_norm": 0.251378737609047,
+      "learning_rate": 8.884898703847544e-07,
+      "loss": 0.434,
+      "step": 1522
+    },
+    {
+      "epoch": 4.695183044315993,
+      "grad_norm": 0.08651539819239572,
+      "learning_rate": 8.70515659618123e-07,
+      "loss": 0.4395,
+      "step": 1523
+    },
+    {
+      "epoch": 4.6982658959537575,
+      "grad_norm": 0.08382155127994485,
+      "learning_rate": 8.527231211262355e-07,
+      "loss": 0.4346,
+      "step": 1524
+    },
+    {
+      "epoch": 4.701348747591522,
+      "grad_norm": 0.08399402892535139,
+      "learning_rate": 8.351123375171455e-07,
+      "loss": 0.438,
+      "step": 1525
+    },
+    {
+      "epoch": 4.704431599229287,
+      "grad_norm": 0.09003259922636092,
+      "learning_rate": 8.176833905550663e-07,
+      "loss": 0.4357,
+      "step": 1526
+    },
+    {
+      "epoch": 4.707514450867052,
+      "grad_norm": 0.08842379329318138,
+      "learning_rate": 8.00436361159953e-07,
+      "loss": 0.4367,
+      "step": 1527
+    },
+    {
+      "epoch": 4.710597302504817,
+      "grad_norm": 0.0880236845833887,
+      "learning_rate": 7.83371329407161e-07,
+      "loss": 0.4382,
+      "step": 1528
+    },
+    {
+      "epoch": 4.7136801541425815,
+      "grad_norm": 0.08572111299033434,
+      "learning_rate": 7.664883745270458e-07,
+      "loss": 0.436,
+      "step": 1529
+    },
+    {
+      "epoch": 4.716763005780347,
+      "grad_norm": 0.08307990367989625,
+      "learning_rate": 7.497875749046124e-07,
+      "loss": 0.4357,
+      "step": 1530
+    },
+    {
+      "epoch": 4.719845857418112,
+      "grad_norm": 0.08392771823337898,
+      "learning_rate": 7.332690080791471e-07,
+      "loss": 0.4365,
+      "step": 1531
+    },
+    {
+      "epoch": 4.722928709055877,
+      "grad_norm": 0.09429708816334033,
+      "learning_rate": 7.169327507438572e-07,
+      "loss": 0.4385,
+      "step": 1532
+    },
+    {
+      "epoch": 4.7260115606936415,
+      "grad_norm": 0.08676649271242363,
+      "learning_rate": 7.007788787455117e-07,
+      "loss": 0.435,
+      "step": 1533
+    },
+    {
+      "epoch": 4.729094412331406,
+      "grad_norm": 0.08308085149438327,
+      "learning_rate": 6.848074670840943e-07,
+      "loss": 0.4365,
+      "step": 1534
+    },
+    {
+      "epoch": 4.732177263969172,
+      "grad_norm": 0.0875675958840115,
+      "learning_rate": 6.690185899124579e-07,
+      "loss": 0.4341,
+      "step": 1535
+    },
+    {
+      "epoch": 4.735260115606937,
+      "grad_norm": 0.08311011794502618,
+      "learning_rate": 6.534123205359555e-07,
+      "loss": 0.4371,
+      "step": 1536
+    },
+    {
+      "epoch": 4.7383429672447015,
+      "grad_norm": 0.08988315522511615,
+      "learning_rate": 6.379887314121514e-07,
+      "loss": 0.4355,
+      "step": 1537
+    },
+    {
+      "epoch": 4.741425818882466,
+      "grad_norm": 0.08632686486107095,
+      "learning_rate": 6.227478941504217e-07,
+      "loss": 0.4409,
+      "step": 1538
+    },
+    {
+      "epoch": 4.744508670520231,
+      "grad_norm": 0.08332899327627469,
+      "learning_rate": 6.076898795116792e-07,
+      "loss": 0.4393,
+      "step": 1539
+    },
+    {
+      "epoch": 4.747591522157996,
+      "grad_norm": 0.08265474203970438,
+      "learning_rate": 5.928147574080001e-07,
+      "loss": 0.4382,
+      "step": 1540
+    },
+    {
+      "epoch": 4.750674373795761,
+      "grad_norm": 0.08784900025025266,
+      "learning_rate": 5.781225969023263e-07,
+      "loss": 0.4374,
+      "step": 1541
+    },
+    {
+      "epoch": 4.753757225433526,
+      "grad_norm": 0.08269780826965292,
+      "learning_rate": 5.636134662081416e-07,
+      "loss": 0.4351,
+      "step": 1542
+    },
+    {
+      "epoch": 4.756840077071291,
+      "grad_norm": 0.08305113967679523,
+      "learning_rate": 5.492874326891384e-07,
+      "loss": 0.4333,
+      "step": 1543
+    },
+    {
+      "epoch": 4.759922928709056,
+      "grad_norm": 0.08772066079329303,
+      "learning_rate": 5.351445628589203e-07,
+      "loss": 0.4388,
+      "step": 1544
+    },
+    {
+      "epoch": 4.763005780346821,
+      "grad_norm": 0.08304807879611681,
+      "learning_rate": 5.211849223806997e-07,
+      "loss": 0.4334,
+      "step": 1545
+    },
+    {
+      "epoch": 4.766088631984585,
+      "grad_norm": 0.09075436377566874,
+      "learning_rate": 5.074085760669656e-07,
+      "loss": 0.4396,
+      "step": 1546
+    },
+    {
+      "epoch": 4.769171483622351,
+      "grad_norm": 0.0910202164296961,
+      "learning_rate": 4.938155878792161e-07,
+      "loss": 0.4358,
+      "step": 1547
+    },
+    {
+      "epoch": 4.772254335260116,
+      "grad_norm": 0.08827541817031775,
+      "learning_rate": 4.804060209276396e-07,
+      "loss": 0.4395,
+      "step": 1548
+    },
+    {
+      "epoch": 4.775337186897881,
+      "grad_norm": 0.09709280833147413,
+      "learning_rate": 4.6717993747082524e-07,
+      "loss": 0.4347,
+      "step": 1549
+    },
+    {
+      "epoch": 4.7784200385356455,
+      "grad_norm": 0.08440596872431598,
+      "learning_rate": 4.5413739891547517e-07,
+      "loss": 0.4387,
+      "step": 1550
+    },
+    {
+      "epoch": 4.78150289017341,
+      "grad_norm": 0.08363374341871826,
+      "learning_rate": 4.412784658161151e-07,
+      "loss": 0.4402,
+      "step": 1551
+    },
+    {
+      "epoch": 4.784585741811175,
+      "grad_norm": 0.0809317188638219,
+      "learning_rate": 4.2860319787484174e-07,
+      "loss": 0.438,
+      "step": 1552
+    },
+    {
+      "epoch": 4.78766859344894,
+      "grad_norm": 0.08121831777790932,
+      "learning_rate": 4.161116539409893e-07,
+      "loss": 0.438,
+      "step": 1553
+    },
+    {
+      "epoch": 4.7907514450867055,
+      "grad_norm": 0.0880895045672302,
+      "learning_rate": 4.0380389201092105e-07,
+      "loss": 0.4374,
+      "step": 1554
+    },
+    {
+      "epoch": 4.79383429672447,
+      "grad_norm": 0.08356116143771772,
+      "learning_rate": 3.9167996922770513e-07,
+      "loss": 0.4385,
+      "step": 1555
+    },
+    {
+      "epoch": 4.796917148362235,
+      "grad_norm": 0.08344910559971451,
+      "learning_rate": 3.7973994188087894e-07,
+      "loss": 0.4389,
+      "step": 1556
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.08950993914377013,
+      "learning_rate": 3.679838654061874e-07,
+      "loss": 0.4398,
+      "step": 1557
+    },
+    {
+      "epoch": 4.803082851637765,
+      "grad_norm": 0.08542262245467172,
+      "learning_rate": 3.5641179438531624e-07,
+      "loss": 0.4415,
+      "step": 1558
+    },
+    {
+      "epoch": 4.80616570327553,
+      "grad_norm": 0.08497388018090479,
+      "learning_rate": 3.450237825456304e-07,
+      "loss": 0.4416,
+      "step": 1559
+    },
+    {
+      "epoch": 4.809248554913295,
+      "grad_norm": 0.08351462868824487,
+      "learning_rate": 3.338198827599559e-07,
+      "loss": 0.4387,
+      "step": 1560
+    },
+    {
+      "epoch": 4.81233140655106,
+      "grad_norm": 0.08772710595626439,
+      "learning_rate": 3.22800147046296e-07,
+      "loss": 0.4421,
+      "step": 1561
+    },
+    {
+      "epoch": 4.815414258188825,
+      "grad_norm": 0.08848355515083191,
+      "learning_rate": 3.119646265676224e-07,
+      "loss": 0.4356,
+      "step": 1562
+    },
+    {
+      "epoch": 4.818497109826589,
+      "grad_norm": 0.0852812752075134,
+      "learning_rate": 3.013133716316174e-07,
+      "loss": 0.4373,
+      "step": 1563
+    },
+    {
+      "epoch": 4.821579961464354,
+      "grad_norm": 0.08552228730185174,
+      "learning_rate": 2.908464316904391e-07,
+      "loss": 0.4366,
+      "step": 1564
+    },
+    {
+      "epoch": 4.824662813102119,
+      "grad_norm": 0.0860392533681976,
+      "learning_rate": 2.805638553405121e-07,
+      "loss": 0.4364,
+      "step": 1565
+    },
+    {
+      "epoch": 4.827745664739885,
+      "grad_norm": 0.08414073695623006,
+      "learning_rate": 2.704656903222791e-07,
+      "loss": 0.4369,
+      "step": 1566
+    },
+    {
+      "epoch": 4.830828516377649,
+      "grad_norm": 0.0833451499983833,
+      "learning_rate": 2.605519835199966e-07,
+      "loss": 0.4389,
+      "step": 1567
+    },
+    {
+      "epoch": 4.833911368015414,
+      "grad_norm": 0.0869990470776574,
+      "learning_rate": 2.508227809614994e-07,
+      "loss": 0.4367,
+      "step": 1568
+    },
+    {
+      "epoch": 4.836994219653179,
+      "grad_norm": 0.0800086392869368,
+      "learning_rate": 2.4127812781800543e-07,
+      "loss": 0.4353,
+      "step": 1569
+    },
+    {
+      "epoch": 4.840077071290944,
+      "grad_norm": 0.0829961020150652,
+      "learning_rate": 2.3191806840389795e-07,
+      "loss": 0.4355,
+      "step": 1570
+    },
+    {
+      "epoch": 4.8431599229287094,
+      "grad_norm": 0.08006058023162158,
+      "learning_rate": 2.2274264617651255e-07,
+      "loss": 0.4357,
+      "step": 1571
+    },
+    {
+      "epoch": 4.846242774566474,
+      "grad_norm": 0.08440440716893022,
+      "learning_rate": 2.1375190373594147e-07,
+      "loss": 0.4353,
+      "step": 1572
+    },
+    {
+      "epoch": 4.849325626204239,
+      "grad_norm": 0.08230957454621195,
+      "learning_rate": 2.0494588282485184e-07,
+      "loss": 0.4356,
+      "step": 1573
+    },
+    {
+      "epoch": 4.852408477842004,
+      "grad_norm": 0.0825814545043019,
+      "learning_rate": 1.9632462432825906e-07,
+      "loss": 0.4402,
+      "step": 1574
+    },
+    {
+      "epoch": 4.855491329479769,
+      "grad_norm": 0.07960269927153034,
+      "learning_rate": 1.8788816827336686e-07,
+      "loss": 0.4389,
+      "step": 1575
+    },
+    {
+      "epoch": 4.858574181117533,
+      "grad_norm": 0.08382296590771703,
+      "learning_rate": 1.7963655382936318e-07,
+      "loss": 0.4356,
+      "step": 1576
+    },
+    {
+      "epoch": 4.861657032755299,
+      "grad_norm": 0.08131641282529142,
+      "learning_rate": 1.7156981930723793e-07,
+      "loss": 0.4354,
+      "step": 1577
+    },
+    {
+      "epoch": 4.864739884393064,
+      "grad_norm": 0.08252787395396438,
+      "learning_rate": 1.6368800215962766e-07,
+      "loss": 0.4369,
+      "step": 1578
+    },
+    {
+      "epoch": 4.867822736030829,
+      "grad_norm": 0.0840846984505899,
+      "learning_rate": 1.5599113898061124e-07,
+      "loss": 0.4394,
+      "step": 1579
+    },
+    {
+      "epoch": 4.870905587668593,
+      "grad_norm": 0.08058386160185066,
+      "learning_rate": 1.4847926550556336e-07,
+      "loss": 0.4374,
+      "step": 1580
+    },
+    {
+      "epoch": 4.873988439306358,
+      "grad_norm": 0.08138706794037397,
+      "learning_rate": 1.4115241661097235e-07,
+      "loss": 0.433,
+      "step": 1581
+    },
+    {
+      "epoch": 4.877071290944123,
+      "grad_norm": 0.08839600750816769,
+      "learning_rate": 1.3401062631429373e-07,
+      "loss": 0.4388,
+      "step": 1582
+    },
+    {
+      "epoch": 4.880154142581889,
+      "grad_norm": 0.08561559096278952,
+      "learning_rate": 1.27053927773777e-07,
+      "loss": 0.4408,
+      "step": 1583
+    },
+    {
+      "epoch": 4.883236994219653,
+      "grad_norm": 0.0823886578375572,
+      "learning_rate": 1.2028235328831906e-07,
+      "loss": 0.4365,
+      "step": 1584
+    },
+    {
+      "epoch": 4.886319845857418,
+      "grad_norm": 0.08125254855231347,
+      "learning_rate": 1.1369593429731318e-07,
+      "loss": 0.4383,
+      "step": 1585
+    },
+    {
+      "epoch": 4.889402697495183,
+      "grad_norm": 0.08424024217952836,
+      "learning_rate": 1.072947013805159e-07,
+      "loss": 0.438,
+      "step": 1586
+    },
+    {
+      "epoch": 4.892485549132948,
+      "grad_norm": 0.08066892423964714,
+      "learning_rate": 1.0107868425786926e-07,
+      "loss": 0.438,
+      "step": 1587
+    },
+    {
+      "epoch": 4.8955684007707125,
+      "grad_norm": 0.08219408326938446,
+      "learning_rate": 9.504791178941209e-08,
+      "loss": 0.4398,
+      "step": 1588
+    },
+    {
+      "epoch": 4.898651252408478,
+      "grad_norm": 0.08589218023326195,
+      "learning_rate": 8.92024119751067e-08,
+      "loss": 0.4369,
+      "step": 1589
+    },
+    {
+      "epoch": 4.901734104046243,
+      "grad_norm": 0.081447283311268,
+      "learning_rate": 8.354221195471912e-08,
+      "loss": 0.4375,
+      "step": 1590
+    },
+    {
+      "epoch": 4.904816955684008,
+      "grad_norm": 0.08509320563460243,
+      "learning_rate": 7.806733800769906e-08,
+      "loss": 0.4383,
+      "step": 1591
+    },
+    {
+      "epoch": 4.907899807321773,
+      "grad_norm": 0.08074788055127495,
+      "learning_rate": 7.277781555306895e-08,
+      "loss": 0.4342,
+      "step": 1592
+    },
+    {
+      "epoch": 4.910982658959537,
+      "grad_norm": 0.08079733820460183,
+      "learning_rate": 6.767366914927298e-08,
+      "loss": 0.438,
+      "step": 1593
+    },
+    {
+      "epoch": 4.914065510597302,
+      "grad_norm": 0.07925496888276341,
+      "learning_rate": 6.275492249409265e-08,
+      "loss": 0.4379,
+      "step": 1594
+    },
+    {
+      "epoch": 4.917148362235068,
+      "grad_norm": 0.08226948438020723,
+      "learning_rate": 5.8021598424522483e-08,
+      "loss": 0.4382,
+      "step": 1595
+    },
+    {
+      "epoch": 4.920231213872833,
+      "grad_norm": 0.08232700981955202,
+      "learning_rate": 5.34737189166723e-08,
+      "loss": 0.4403,
+      "step": 1596
+    },
+    {
+      "epoch": 4.923314065510597,
+      "grad_norm": 0.07991121522338393,
+      "learning_rate": 4.9111305085656204e-08,
+      "loss": 0.4419,
+      "step": 1597
+    },
+    {
+      "epoch": 4.926396917148362,
+      "grad_norm": 0.08113247408332278,
+      "learning_rate": 4.493437718550375e-08,
+      "loss": 0.4374,
+      "step": 1598
+    },
+    {
+      "epoch": 4.929479768786127,
+      "grad_norm": 0.08233516296821973,
+      "learning_rate": 4.094295460905784e-08,
+      "loss": 0.4381,
+      "step": 1599
+    },
+    {
+      "epoch": 4.932562620423893,
+      "grad_norm": 0.08179952941671956,
+      "learning_rate": 3.7137055887890294e-08,
+      "loss": 0.4377,
+      "step": 1600
+    },
+    {
+      "epoch": 4.935645472061657,
+      "grad_norm": 0.08322558616105742,
+      "learning_rate": 3.351669869221308e-08,
+      "loss": 0.4369,
+      "step": 1601
+    },
+    {
+      "epoch": 4.938728323699422,
+      "grad_norm": 0.08221662807478401,
+      "learning_rate": 3.0081899830798345e-08,
+      "loss": 0.4375,
+      "step": 1602
+    },
+    {
+      "epoch": 4.941811175337187,
+      "grad_norm": 0.08397725385709745,
+      "learning_rate": 2.683267525089406e-08,
+      "loss": 0.4404,
+      "step": 1603
+    },
+    {
+      "epoch": 4.944894026974952,
+      "grad_norm": 0.08432100843835037,
+      "learning_rate": 2.3769040038157387e-08,
+      "loss": 0.4402,
+      "step": 1604
+    },
+    {
+      "epoch": 4.9479768786127165,
+      "grad_norm": 0.08399935387842421,
+      "learning_rate": 2.089100841658809e-08,
+      "loss": 0.4356,
+      "step": 1605
+    },
+    {
+      "epoch": 4.951059730250481,
+      "grad_norm": 0.08114538291099105,
+      "learning_rate": 1.8198593748448568e-08,
+      "loss": 0.4377,
+      "step": 1606
+    },
+    {
+      "epoch": 4.954142581888247,
+      "grad_norm": 0.08165399487597627,
+      "learning_rate": 1.5691808534210595e-08,
+      "loss": 0.4337,
+      "step": 1607
+    },
+    {
+      "epoch": 4.957225433526012,
+      "grad_norm": 0.083769651895536,
+      "learning_rate": 1.337066441250201e-08,
+      "loss": 0.4371,
+      "step": 1608
+    },
+    {
+      "epoch": 4.9603082851637765,
+      "grad_norm": 0.08280256101776927,
+      "learning_rate": 1.1235172160044549e-08,
+      "loss": 0.4318,
+      "step": 1609
+    },
+    {
+      "epoch": 4.963391136801541,
+      "grad_norm": 0.0799765532014931,
+      "learning_rate": 9.285341691600558e-09,
+      "loss": 0.4384,
+      "step": 1610
+    },
+    {
+      "epoch": 4.966473988439306,
+      "grad_norm": 0.08068921429375234,
+      "learning_rate": 7.521182059946342e-09,
+      "loss": 0.4378,
+      "step": 1611
+    },
+    {
+      "epoch": 4.969556840077072,
+      "grad_norm": 0.07966770614287762,
+      "learning_rate": 5.942701455796673e-09,
+      "loss": 0.4391,
+      "step": 1612
+    },
+    {
+      "epoch": 4.9726396917148366,
+      "grad_norm": 0.08203012049591347,
+      "learning_rate": 4.549907207804794e-09,
+      "loss": 0.4376,
+      "step": 1613
+    },
+    {
+      "epoch": 4.975722543352601,
+      "grad_norm": 0.08028826748789267,
+      "learning_rate": 3.3428057825091177e-09,
+      "loss": 0.4392,
+      "step": 1614
+    },
+    {
+      "epoch": 4.978805394990366,
+      "grad_norm": 0.08225551366500912,
+      "learning_rate": 2.3214027842888287e-09,
+      "loss": 0.4387,
+      "step": 1615
+    },
+    {
+      "epoch": 4.981888246628131,
+      "grad_norm": 0.07989935782081639,
+      "learning_rate": 1.485702955363877e-09,
+      "loss": 0.439,
+      "step": 1616
+    },
+    {
+      "epoch": 4.984971098265896,
+      "grad_norm": 0.0824757445942737,
+      "learning_rate": 8.357101757683339e-10,
+      "loss": 0.435,
+      "step": 1617
+    },
+    {
+      "epoch": 4.9880539499036605,
+      "grad_norm": 0.08106848336497609,
+      "learning_rate": 3.7142746331930714e-10,
+      "loss": 0.4391,
+      "step": 1618
+    },
+    {
+      "epoch": 4.991136801541426,
+      "grad_norm": 0.07969019966943018,
+      "learning_rate": 9.285697361249845e-11,
+      "loss": 0.4358,
+      "step": 1619
+    },
+    {
+      "epoch": 4.994219653179191,
+      "grad_norm": 0.08398435288446787,
+      "learning_rate": 0.0,
+      "loss": 0.4371,
+      "step": 1620
+    },
+    {
+      "epoch": 4.994219653179191,
+      "step": 1620,
+      "total_flos": 4.168742448987636e+19,
+      "train_loss": 0.5089499812619186,
+      "train_runtime": 355375.6593,
+      "train_samples_per_second": 2.337,
+      "train_steps_per_second": 0.005
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1620,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.168742448987636e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}